paper_id uint32 0 4.07k | title stringlengths 8 154 | authors listlengths 1 40 | cvf_url stringlengths 86 196 | pdf_url stringlengths 87 197 | supp_url stringlengths 98 147 ⌀ | arxiv_id stringlengths 10 10 ⌀ | arxiv_id_source stringclasses 4
values | bibtex large_stringlengths 308 1.06k | abstract large_stringlengths 562 2.77k | embedding listlengths 768 768 |
|---|---|---|---|---|---|---|---|---|---|---|
100 | Towards Robust Vision Transformers: Path Dependency Analysis and a Simple Two-Stage Adversarial Training | [
"Seongmin Kim",
"Byung Cheol Song"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Kim_Towards_Robust_Vision_Transformers_Path_Dependency_Analysis_and_a_Simple_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Kim_Towards_Robust_Vision_Transformers_Path_Dependency_Analysis_and_a_Simple_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Kim_Towards_Robust_Vision_CVPR_2026_supplemental.pdf | null | null | @InProceedings{Kim_2026_CVPR,
author = {Kim, Seongmin and Song, Byung Cheol},
title = {Towards Robust Vision Transformers: Path Dependency Analysis and a Simple Two-Stage Adversarial Training},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
... | The Vision Transformer (ViT) has surpassed Convolutional Neural Networks (CNNs) in performance, becoming the de facto architecture in modern computer vision. However, despite its superior representational capacity, research on the adversarial robustness of ViTs remains limited, with most studies still biased toward CNN... | [
0.021229006350040436,
-0.021151961758732796,
0.010618451982736588,
0.044249240309000015,
0.014320207759737968,
0.021343499422073364,
0.0391932912170887,
0.004113059025257826,
-0.0105535127222538,
-0.04752066731452942,
-0.022545618936419487,
0.031153785064816475,
-0.069023497402668,
0.00118... |
101 | First Frame Is the Place to Go for Video Content Customization | [
"Jingxi Chen",
"Zongxia Li",
"Zhichao Liu",
"Guangyao Shi",
"Xiyang Wu",
"Fuxiao Liu",
"Cornelia Fermüller",
"Brandon Y. Feng",
"Yiannis Aloimonos"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Chen_First_Frame_Is_the_Place_to_Go_for_Video_Content_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Chen_First_Frame_Is_the_Place_to_Go_for_Video_Content_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Chen_First_Frame_Is_CVPR_2026_supplemental.pdf | 2511.15700 | cvf | @InProceedings{Chen_2026_CVPR,
author = {Chen, Jingxi and Li, Zongxia and Liu, Zhichao and Shi, Guangyao and Wu, Xiyang and Liu, Fuxiao and Ferm\"uller, Cornelia and Feng, Brandon Y. and Aloimonos, Yiannis},
title = {First Frame Is the Place to Go for Video Content Customization},
booktitle = {Procee... | What role does the first frame play in video generation models? Traditionally, it's viewed as the spatial-temporal starting point of a video, merely a seed for subsequent animation. In this work, we reveal a fundamentally different perspective: video models implicitly treat the first frame as a conceptual memory buffer... | [
0.031109942123293877,
0.002147297840565443,
0.015833957120776176,
0.041221242398023605,
0.05486442521214485,
0.0007732658414170146,
0.015322878956794739,
0.03892619162797928,
-0.0388481430709362,
-0.061773184686899185,
-0.04796244576573372,
-0.01364970300346613,
-0.045762911438941956,
0.01... |
102 | High-Quality and Efficient Turbulence Mitigation with Events | [
"Xiaoran Zhang",
"Jian Ding",
"Yuxing Duan",
"Haoyue Liu",
"Gang Chen",
"Yi Chang",
"Luxin Yan"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Zhang_High-Quality_and_Efficient_Turbulence_Mitigation_with_Events_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Zhang_High-Quality_and_Efficient_Turbulence_Mitigation_with_Events_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zhang_High-Quality_and_Efficient_CVPR_2026_supplemental.zip | 2603.20708 | cvf | @InProceedings{Zhang_2026_CVPR,
author = {Zhang, Xiaoran and Ding, Jian and Duan, Yuxing and Liu, Haoyue and Chen, Gang and Chang, Yi and Yan, Luxin},
title = {High-Quality and Efficient Turbulence Mitigation with Events},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pat... | Turbulence mitigation (TM) is highly ill-posed due to the stochastic nature of atmospheric turbulence. Most methods rely on multiple frames recorded by conventional cameras to capture stable patterns in natural scenarios. However, they inevitably suffer from a trade-off between accuracy and efficiency: more frames enha... | [
0.030448555946350098,
0.012256581336259842,
0.0009161669295281172,
0.05606313794851303,
0.03919240087270737,
0.012425211258232594,
0.012428835965692997,
0.014644897542893887,
-0.06365236639976501,
-0.059344299137592316,
-0.020449522882699966,
-0.016004949808120728,
-0.04065963253378868,
-0... |
103 | HeroGS: Hierarchical Guidance for Robust 3D Gaussian Splatting under Sparse Views | [
"Jiashu Li",
"Xumeng Han",
"Zhaoyang Wei",
"Zipeng Wang",
"Kuiran Wang",
"Guorong Li",
"Zhenjun Han",
"Jianbin Jiao"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Li_HeroGS_Hierarchical_Guidance_for_Robust_3D_Gaussian_Splatting_under_Sparse_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Li_HeroGS_Hierarchical_Guidance_for_Robust_3D_Gaussian_Splatting_under_Sparse_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Li_HeroGS_Hierarchical_Guidance_CVPR_2026_supplemental.pdf | 2603.01099 | cvf | @InProceedings{Li_2026_CVPR,
author = {Li, Jiashu and Han, Xumeng and Wei, Zhaoyang and Wang, Zipeng and Wang, Kuiran and Li, Guorong and Han, Zhenjun and Jiao, Jianbin},
title = {HeroGS: Hierarchical Guidance for Robust 3D Gaussian Splatting under Sparse Views},
booktitle = {Proceedings of the IEEE/... | 3D Gaussian Splatting (3DGS) has recently emerged as a promising approach in novel view synthesis, combining photorealistic rendering with real-time efficiency. However, its success heavily relies on dense camera coverage; under sparse-view conditions, insufficient supervision leads to irregular Gaussian distributions-... | [
0.006732921116054058,
0.008975805714726448,
0.030251622200012207,
0.03136402741074562,
0.014071108773350716,
0.027607591822743416,
0.010560398921370506,
0.009084992110729218,
-0.045785121619701385,
-0.058693669736385345,
-0.009418907575309277,
-0.0368928462266922,
-0.05694286525249481,
0.0... |
104 | SEA-Vision: A Multilingual Benchmark for Comprehensive Document and Scene Text Understanding in Southeast Asia | [
"Pengfei Yue",
"Xingran Zhao",
"Juntao Chen",
"Peng Hou",
"Wang Longchao",
"Jianghang Lin",
"Shengchuan Zhang",
"Anxiang Zeng",
"Liujuan Cao"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Yue_SEA-Vision_A_Multilingual_Benchmark_for_Comprehensive_Document_and_Scene_Text_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Yue_SEA-Vision_A_Multilingual_Benchmark_for_Comprehensive_Document_and_Scene_Text_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Yue_SEA-Vision_A_Multilingual_CVPR_2026_supplemental.pdf | 2603.15409 | cvf | @InProceedings{Yue_2026_CVPR,
author = {Yue, Pengfei and Zhao, Xingran and Chen, Juntao and Hou, Peng and Longchao, Wang and Lin, Jianghang and Zhang, Shengchuan and Zeng, Anxiang and Cao, Liujuan},
title = {SEA-Vision: A Multilingual Benchmark for Comprehensive Document and Scene Text Understanding in S... | Multilingual document and scene text understanding plays an important role in applications such as search, finance, and public services. However, most existing benchmarks focus on high-resource languages and fail to evaluate models in realistic multilingual environments. In Southeast Asia, the diversity of languages, c... | [
-0.010884860530495644,
0.007055943366140127,
-0.014838876202702522,
0.047815099358558655,
0.002837520558387041,
0.00871484074741602,
-0.005041626747697592,
0.047498226165771484,
-0.03250398859381676,
0.004578393884003162,
-0.01856391504406929,
0.04243910685181618,
-0.05737631767988205,
-0.... |
105 | VITAL: Vision-Encoder-centered Pre-training for LMMs in Visual Quality Assessment | [
"Ziheng Jia",
"Linhan Cao",
"Jinliang Han",
"Zicheng Zhang",
"Jiaying Qian",
"Jiarui Wang",
"Zijian Chen",
"Guangtao Zhai",
"Xiongkuo Min"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Jia_VITAL_Vision-Encoder-centered_Pre-training_for_LMMs_in_Visual_Quality_Assessment_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Jia_VITAL_Vision-Encoder-centered_Pre-training_for_LMMs_in_Visual_Quality_Assessment_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Jia_VITAL_Vision-Encoder-centered_Pre-training_CVPR_2026_supplemental.zip | 2511.17962 | cvf | @InProceedings{Jia_2026_CVPR,
author = {Jia, Ziheng and Cao, Linhan and Han, Jinliang and Zhang, Zicheng and Qian, Jiaying and Wang, Jiarui and Chen, Zijian and Zhai, Guangtao and Min, Xiongkuo},
title = {VITAL: Vision-Encoder-centered Pre-training for LMMs in Visual Quality Assessment},
booktitle = ... | Developing a robust visual quality assessment (VQualA) large multi-modal model (LMM) requires achieving versatility, powerfulness, and transferability. However, existing VQualA LMMs typically focus on a single task and rely on full-parameter fine-tuning, which makes them prone to overfitting on specific modalities or t... | [
0.04820045083761215,
-0.0029130030889064074,
0.02443857491016388,
0.03511703759431839,
0.031503111124038696,
0.05472537502646446,
0.02649097889661789,
0.01847177743911743,
-0.05634467676281929,
-0.04241073131561279,
-0.04709574952721596,
0.02256077155470848,
-0.048850156366825104,
0.013406... |
106 | Jailbreaking Vision-Language Models via Dissonance-Guided Suffix Optimization and Image-Phrase Injection | [
"Jiacheng Pi",
"Zhiguo Yang",
"Xingxing Huang",
"Dongsheng Xu",
"Ruizhi Zhong",
"Wenjie Ruan"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Pi_Jailbreaking_Vision-Language_Models_via_Dissonance-Guided_Suffix_Optimization_and_Image-Phrase_Injection_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Pi_Jailbreaking_Vision-Language_Models_via_Dissonance-Guided_Suffix_Optimization_and_Image-Phrase_Injection_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Pi_Jailbreaking_Vision-Language_Models_CVPR_2026_supplemental.pdf | null | null | @InProceedings{Pi_2026_CVPR,
author = {Pi, Jiacheng and Yang, Zhiguo and Huang, Xingxing and Xu, Dongsheng and Zhong, Ruizhi and Ruan, Wenjie},
title = {Jailbreaking Vision-Language Models via Dissonance-Guided Suffix Optimization and Image-Phrase Injection},
booktitle = {Proceedings of the IEEE/CVF ... | The integration of vision and language in Vision-Language Models (VLMs), while enabling multimodal capabilities, inherently expands their attack surface. Among existing white-box jailbreak methods, suffix-optimization-based approaches often rely on gradient approximations over discrete token spaces, yielding insufficie... | [
0.0042786793783307076,
-0.010278533212840557,
-0.0007027725805528462,
0.05574986711144447,
0.01093941181898117,
0.05838434025645256,
0.05954483523964882,
-0.0018292629392817616,
-0.040422193706035614,
-0.01647360809147358,
-0.026268158107995987,
0.008748563937842846,
-0.09861699491739273,
... |
107 | PDD: Manifold-Prior Diverse Distillation for Medical Anomaly Detection | [
"Xijun Lu",
"Hongying Liu",
"Fanhua Shang",
"Yanming Hui",
"Liang Wan"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Lu_PDD_Manifold-Prior_Diverse_Distillation_for_Medical_Anomaly_Detection_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Lu_PDD_Manifold-Prior_Diverse_Distillation_for_Medical_Anomaly_Detection_CVPR_2026_paper.pdf | null | 2603.07142 | cvf | @InProceedings{Lu_2026_CVPR,
author = {Lu, Xijun and Liu, Hongying and Shang, Fanhua and Hui, Yanming and Wan, Liang},
title = {PDD: Manifold-Prior Diverse Distillation for Medical Anomaly Detection},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR... | Medical image anomaly detection faces unique challenges due to subtle, heterogeneous anomalies embedded in complex anatomical structures. Through systematic Grad-CAM analysis, we reveal that discriminative activation maps fail on medical data, unlike their success on industrial datasets, motivating the need for manifol... | [
-0.01705801859498024,
-0.009811795316636562,
-0.03335815295577049,
0.06021399796009064,
0.033213257789611816,
0.03405245766043663,
0.03190493956208229,
-0.029165685176849365,
-0.04125852510333061,
-0.05161282792687416,
0.012043365277349949,
0.002170708728954196,
-0.057598039507865906,
0.02... |
108 | Bidirectional Query-Driven Generation of Parametric CAD Sketch | [
"Yang Liu",
"Daxuan Ren",
"Yijie Ding",
"Jianmin Zheng",
"Fang Deng"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Liu_Bidirectional_Query-Driven_Generation_of_Parametric_CAD_Sketch_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Liu_Bidirectional_Query-Driven_Generation_of_Parametric_CAD_Sketch_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Liu_Bidirectional_Query-Driven_Generation_CVPR_2026_supplemental.pdf | null | null | @InProceedings{Liu_2026_CVPR,
author = {Liu, Yang and Ren, Daxuan and Ding, Yijie and Zheng, Jianmin and Deng, Fang},
title = {Bidirectional Query-Driven Generation of Parametric CAD Sketch},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
m... | Learning-based CAD modeling shows great promise in automating parametric design, yet existing approaches often overlook the incremental and state-dependent nature of sketch construction. We present CADSketcher, a query-driven bidirectional framework for completing partial parametric sketches by internalizing the non-li... | [
-0.004572725854814053,
-0.010885190218687057,
-0.019376853480935097,
0.03780801221728325,
0.017810216173529625,
0.05125704035162926,
0.010934311896562576,
0.00831194594502449,
-0.003724837675690651,
-0.08145514875650406,
-0.02116077020764351,
-0.021701868623495102,
-0.03673297539353371,
0.... |
109 | Illustrator's Depth: Monocular Layer Index Prediction for Image Decomposition | [
"Nissim Maruani",
"Peiying Zhang",
"Siddhartha Chaudhuri",
"Matthew Fisher",
"Nanxuan Zhao",
"Vladimir G. Kim",
"Pierre Alliez",
"Mathieu Desbrun",
"Wang Yifan"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Maruani_Illustrators_Depth_Monocular_Layer_Index_Prediction_for_Image_Decomposition_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Maruani_Illustrators_Depth_Monocular_Layer_Index_Prediction_for_Image_Decomposition_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Maruani_Illustrators_Depth_Monocular_CVPR_2026_supplemental.pdf | 2511.17454 | title_snapshot | @InProceedings{Maruani_2026_CVPR,
author = {Maruani, Nissim and Zhang, Peiying and Chaudhuri, Siddhartha and Fisher, Matthew and Zhao, Nanxuan and Kim, Vladimir G. and Alliez, Pierre and Desbrun, Mathieu and Yifan, Wang},
title = {Illustrator's Depth: Monocular Layer Index Prediction for Image Decomposit... | We introduce Illustrator's Depth, a novel definition of depth that addresses a key challenge in digital content creation: decomposing flat images into editable, ordered layers. Inspired by an artist's compositional process, illustrator's depth infers a layer index for each pixel, forming an interpretable image decompos... | [
-0.01219213381409645,
-0.013337656855583191,
0.011403745971620083,
0.016649683937430382,
0.025112224742770195,
0.022333918139338493,
0.01601462811231613,
0.008169209584593773,
-0.03828587010502815,
-0.0717538595199585,
-0.03558016195893288,
0.005722091067582369,
-0.05384982004761696,
0.035... |
110 | On the Role of Temporal Granularity in the Robustness of Spiking Neural Networks | [
"Mengting Xu",
"Shi Gu",
"Peng Lin",
"De Ma",
"Huajin Tang",
"Qian Zheng",
"Gang Pan"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Xu_On_the_Role_of_Temporal_Granularity_in_the_Robustness_of_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Xu_On_the_Role_of_Temporal_Granularity_in_the_Robustness_of_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Xu_On_the_Role_CVPR_2026_supplemental.pdf | null | null | @InProceedings{Xu_2026_CVPR,
author = {Xu, Mengting and Gu, Shi and Lin, Peng and Ma, De and Tang, Huajin and Zheng, Qian and Pan, Gang},
title = {On the Role of Temporal Granularity in the Robustness of Spiking Neural Networks},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision ... | As the third generation of neural networks, Spiking Neural Networks (SNNs) have demonstrated remarkable potential across diverse applications owing to their unique temporal dynamics. In recent years, analyzing the robustness of SNNs from a temporal perspective has become an emerging research focus. However, most existi... | [
-0.011887090280652046,
-0.012385000474750996,
0.007041115313768387,
0.05387410894036293,
0.025760900229215622,
0.004401342011988163,
0.04777507483959198,
0.014964327216148376,
-0.046778324991464615,
-0.026939302682876587,
0.031303875148296356,
-0.0401507131755352,
-0.041505053639411926,
0.... |
111 | SRGCD: Stability-Driven Region Growth Framework for 3D Change Detection | [
"Yue Wu",
"Tao Peng",
"Yongzhe Yuan",
"Kaiyuan Feng",
"Hao Li",
"Maoguo Gong",
"Qiguang Miao",
"Wenping Ma"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Wu_SRGCD_Stability-Driven_Region_Growth_Framework_for_3D_Change_Detection_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Wu_SRGCD_Stability-Driven_Region_Growth_Framework_for_3D_Change_Detection_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Wu_SRGCD_Stability-Driven_Region_CVPR_2026_supplemental.pdf | null | null | @InProceedings{Wu_2026_CVPR,
author = {Wu, Yue and Peng, Tao and Yuan, Yongzhe and Feng, Kaiyuan and Li, Hao and Gong, Maoguo and Miao, Qiguang and Ma, Wenping},
title = {SRGCD: Stability-Driven Region Growth Framework for 3D Change Detection},
booktitle = {Proceedings of the IEEE/CVF Conference on C... | With the growing accessibility of large-scale 3D point clouds from LiDAR and photogrammetric techniques, 3D change detection (3DCD) has become essential for understanding dynamic scenes. Existing methods typically formulate this as segmentation, treating each point independently for binary classification. This leads to... | [
0.021436195820569992,
-0.01986299827694893,
0.018352555111050606,
0.03537606820464134,
0.035322606563568115,
0.04996158555150032,
0.01485616434365511,
0.014350440353155136,
-0.0500296913087368,
-0.06452330201864243,
-0.0433068722486496,
-0.02551368437707424,
-0.06816437095403671,
0.0111142... |
112 | MoCoDiff: A Controllable Autoregressive Diffusion Model for Expressive Motion Generation | [
"Wenfeng Song",
"Xuehan Wang",
"Shuai Li",
"Yi Chen",
"Yuting Guo",
"Zhenyu Wu",
"Xingliang Jin",
"Chenglizhao Chen",
"Fei Hou",
"Hongyu Wu",
"Aimin Hao"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Song_MoCoDiff_A_Controllable_Autoregressive_Diffusion_Model_for_Expressive_Motion_Generation_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Song_MoCoDiff_A_Controllable_Autoregressive_Diffusion_Model_for_Expressive_Motion_Generation_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Song_MoCoDiff_A_Controllable_CVPR_2026_supplemental.zip | null | null | @InProceedings{Song_2026_CVPR,
author = {Song, Wenfeng and Wang, Xuehan and Li, Shuai and Chen, Yi and Guo, Yuting and Wu, Zhenyu and Jin, Xingliang and Chen, Chenglizhao and Hou, Fei and Wu, Hongyu and Hao, Aimin},
title = {MoCoDiff: A Controllable Autoregressive Diffusion Model for Expressive Motion Ge... | Diffusion-based motion generation has advanced rapidly, but current methods still struggle with long-horizon consistency, style control, and multi-condition guidance. A major reason is the fused-conditioning design, where semantic, stylistic, and temporal signals share a single pathway, causing interference and limitin... | [
-0.028088759630918503,
-0.024389168247580528,
-0.02663996070623398,
0.0464429073035717,
0.05850471556186676,
0.005838178563863039,
0.040762025862932205,
-0.01396797876805067,
-0.01856013759970665,
-0.060693442821502686,
0.023999523371458054,
-0.019863344728946686,
-0.04839424043893814,
-0.... |
113 | SHARP: Short-Window Streaming for Accurate and Robust Prediction in Motion Forecasting | [
"Alexander Prutsch",
"Christian Fruhwirth-Reisinger",
"David Schinagl",
"Horst Possegger"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Prutsch_SHARP_Short-Window_Streaming_for_Accurate_and_Robust_Prediction_in_Motion_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Prutsch_SHARP_Short-Window_Streaming_for_Accurate_and_Robust_Prediction_in_Motion_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Prutsch_SHARP_Short-Window_Streaming_CVPR_2026_supplemental.pdf | 2603.28091 | title_snapshot | @InProceedings{Prutsch_2026_CVPR,
author = {Prutsch, Alexander and Fruhwirth-Reisinger, Christian and Schinagl, David and Possegger, Horst},
title = {SHARP: Short-Window Streaming for Accurate and Robust Prediction in Motion Forecasting},
booktitle = {Proceedings of the IEEE/CVF Conference on Compute... | In dynamic traffic environments, motion forecasting models must be able to accurately estimate future trajectories continuously. Streaming-based methods are a promising solution, but despite recent advances, their performance often degrades when exposed to heterogeneous observation lengths. To address this, we propose ... | [
-0.00369821535423398,
-0.05497679486870766,
0.04764057695865631,
0.035527411848306656,
0.025861350819468498,
0.024812214076519012,
0.03998192027211189,
0.03142189234495163,
-0.04094019904732704,
-0.05677349492907524,
-0.01611683890223503,
0.017912477254867554,
-0.08030306547880173,
-0.0110... |
114 | Dual-Estimator: Decoupling Global and Local Semantic Shift for Drift Compensation in Class-Incremental Learning | [
"Fankang Xu",
"Lu Jin",
"Yanpeng Sun",
"Shiyu Xuan",
"Zechao Li"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Xu_Dual-Estimator_Decoupling_Global_and_Local_Semantic_Shift_for_Drift_Compensation_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Xu_Dual-Estimator_Decoupling_Global_and_Local_Semantic_Shift_for_Drift_Compensation_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Xu_Dual-Estimator_Decoupling_Global_CVPR_2026_supplemental.pdf | null | null | @InProceedings{Xu_2026_CVPR,
author = {Xu, Fankang and Jin, Lu and Sun, Yanpeng and Xuan, Shiyu and Li, Zechao},
title = {Dual-Estimator: Decoupling Global and Local Semantic Shift for Drift Compensation in Class-Incremental Learning},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer V... | Continual Learning (CL) provides an effective paradigm for acquiring new knowledge, and the principle of learning without retaining past samples has led to exemplar-free CL that better matches practical conditions. However, a key challenge is the semantic shift, which requires reliable activation of past class represen... | [
0.015988215804100037,
-0.018405143171548843,
-0.0023935020435601473,
0.0269546527415514,
0.04111999273300171,
0.021704692393541336,
0.019857898354530334,
0.013593369163572788,
-0.00029165635351091623,
-0.030049482360482216,
-0.012183945626020432,
0.016622697934508324,
-0.07272043079137802,
... |
115 | SAM 3D Body: Robust Full-Body Human Mesh Recovery | [
"Xitong Yang",
"Devansh Kukreja",
"Don Pinkus",
"Taosha Fan",
"Jinhyung Park",
"Soyong Shin",
"Jinkun Cao",
"Jia-Wei Liu",
"Nicolás Ugrinovic",
"Anushka Sagar",
"Jitendra Malik",
"Matt Feiszli",
"Piotr Dollár",
"Kris Kitani"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Yang_SAM_3D_Body_Robust_Full-Body_Human_Mesh_Recovery_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Yang_SAM_3D_Body_Robust_Full-Body_Human_Mesh_Recovery_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Yang_SAM_3D_Body_CVPR_2026_supplemental.pdf | 2602.15989 | cvf | @InProceedings{Yang_2026_CVPR,
author = {Yang, Xitong and Kukreja, Devansh and Pinkus, Don and Fan, Taosha and Park, Jinhyung and Shin, Soyong and Cao, Jinkun and Liu, Jia-Wei and Ugrinovic, Nicol\'as and Sagar, Anushka and Malik, Jitendra and Feiszli, Matt and Doll\'ar, Piotr and Kitani, Kris},
title = ... | We introduce SAM 3D Body (3DB), a promptable model for single-image full-body 3D human mesh recovery (HMR) that demonstrates state-of-the-art performance, with strong generalization and consistent accuracy in diverse in-the-wild conditions. 3DB estimates the human pose of the body, feet, and hands. It is the first mode... | [
0.009293075650930405,
-0.007141254376620054,
-0.0248623825609684,
0.022295435890555382,
0.03750767558813095,
0.01113859098404646,
0.04750492796301842,
-0.024063240736722946,
-0.058095209300518036,
-0.06882210820913315,
-0.039839643985033035,
-0.047726958990097046,
-0.07738502323627472,
-0.... |
116 | Boosting Quantitive and Spatial Awareness for Zero-Shot Object Counting | [
"Da Zhang",
"Bingyu Li",
"Feiyu Wang",
"Zhiyuan Zhao",
"Junyu Gao"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Zhang_Boosting_Quantitive_and_Spatial_Awareness_for_Zero-Shot_Object_Counting_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Zhang_Boosting_Quantitive_and_Spatial_Awareness_for_Zero-Shot_Object_Counting_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zhang_Boosting_Quantitive_and_CVPR_2026_supplemental.pdf | 2603.16129 | cvf | @InProceedings{Zhang_2026_CVPR,
author = {Zhang, Da and Li, Bingyu and Wang, Feiyu and Zhao, Zhiyuan and Gao, Junyu},
title = {Boosting Quantitive and Spatial Awareness for Zero-Shot Object Counting},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR... | Zero-shot object counting (ZSOC) aims to enumerate objects of arbitrary categories specified by text descriptions without requiring visual exemplars. However, existing methods often treat counting as a coarse retrieval task, suffering from a lack of fine-grained quantity awareness. Furthermore, they frequently exhibit ... | [
-0.006312601268291473,
-0.021818244829773903,
-0.005229947157204151,
0.019866885617375374,
0.011096351779997349,
0.04959401488304138,
0.014469901099801064,
0.029912158846855164,
-0.05583176761865616,
-0.02600272372364998,
-0.03390306979417801,
0.015026121400296688,
-0.050174131989479065,
-... |
117 | OpenVoxel: Training-Free Grouping and Captioning Voxels for Open-Vocabulary 3D Scene Understanding | [
"Sheng-Yu Huang",
"Jaesung Choe",
"Yu-Chiang Frank Wang",
"Cheng Sun"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Huang_OpenVoxel_Training-Free_Grouping_and_Captioning_Voxels_for_Open-Vocabulary_3D_Scene_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Huang_OpenVoxel_Training-Free_Grouping_and_Captioning_Voxels_for_Open-Vocabulary_3D_Scene_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Huang_OpenVoxel_Training-Free_Grouping_CVPR_2026_supplemental.pdf | 2601.09575 | cvf | @InProceedings{Huang_2026_CVPR,
author = {Huang, Sheng-Yu and Choe, Jaesung and Wang, Yu-Chiang Frank and Sun, Cheng},
title = {OpenVoxel: Training-Free Grouping and Captioning Voxels for Open-Vocabulary 3D Scene Understanding},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision a... | We propose OpenVoxel, a training-free algorithm for grouping and captioning sparse voxels for the open-vocabulary 3D scene understanding tasks. Given the sparse voxel rasterization (SVR) model obtained from multi-view images of a 3D scene, our OpenVoxel is able to produce meaningful groups that describe different objec... | [
0.0030146045610308647,
0.002843153430148959,
0.0501822903752327,
0.0383463054895401,
-0.007085608784109354,
0.018222834914922714,
0.029462488368153572,
0.04166819900274277,
-0.03298654407262802,
-0.01600763201713562,
-0.04485445097088814,
0.017229663208127022,
-0.08225791156291962,
0.01811... |
118 | GuardTrace-VL: Detecting Unsafe Multimodel Reasoning via Iterative Safety Supervision | [
"Yuxiao Xiang",
"Junchi Chen",
"Zhenchao Jin",
"Changtao Miao",
"Haojie Yuan",
"Qi Chu",
"Tao Gong",
"Nenghai Yu"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Xiang_GuardTrace-VL_Detecting_Unsafe_Multimodel_Reasoning_via_Iterative_Safety_Supervision_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Xiang_GuardTrace-VL_Detecting_Unsafe_Multimodel_Reasoning_via_Iterative_Safety_Supervision_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Xiang_GuardTrace-VL_Detecting_Unsafe_CVPR_2026_supplemental.pdf | 2511.20994 | cvf | @InProceedings{Xiang_2026_CVPR,
author = {Xiang, Yuxiao and Chen, Junchi and Jin, Zhenchao and Miao, Changtao and Yuan, Haojie and Chu, Qi and Gong, Tao and Yu, Nenghai},
title = {GuardTrace-VL: Detecting Unsafe Multimodel Reasoning via Iterative Safety Supervision},
booktitle = {Proceedings of the I... | Multimodal large reasoning models (MLRMs) are increasingly deployed for vision-language tasks that produce explicit intermediate rationales. However, reasoning traces can contain unsafe content even when the final answer is non-harmful, creating deployment risks. Existing multimodal safety guards primarily evaluate onl... | [
0.01216849684715271,
0.00793838407844305,
0.009662157855927944,
0.05669613927602768,
0.03803763538599014,
-0.007180975284427404,
0.04157442972064018,
-0.006964109372347593,
-0.03317900002002716,
-0.02512594684958458,
-0.034259747713804245,
0.040360260754823685,
-0.04999052733182907,
-0.005... |
119 | SPAN: Spatial-Projection Alignment for Monocular 3D Object Detection | [
"Yifan Wang",
"Yian Zhao",
"Fanqi Pu",
"Xiaochen Yang",
"Yang Tang",
"Xi Chen",
"Wenming Yang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Wang_SPAN_Spatial-Projection_Alignment_for_Monocular_3D_Object_Detection_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Wang_SPAN_Spatial-Projection_Alignment_for_Monocular_3D_Object_Detection_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Wang_SPAN_Spatial-Projection_Alignment_CVPR_2026_supplemental.pdf | 2511.06702 | cvf | @InProceedings{Wang_2026_CVPR,
author = {Wang, Yifan and Zhao, Yian and Pu, Fanqi and Yang, Xiaochen and Tang, Yang and Chen, Xi and Yang, Wenming},
title = {SPAN: Spatial-Projection Alignment for Monocular 3D Object Detection},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision a... | Existing monocular 3D detectors typically tame the pronounced nonlinear regression of 3D bounding box through decoupled prediction paradigm, which employs multiple branches to estimate geometric center, depth, dimensions, and rotation angle separately.Although this decoupling strategy simplifies the learning process, i... | [
0.04030013456940651,
0.028921624645590782,
0.010906360112130642,
0.025455258786678314,
0.01629527471959591,
0.052436478435993195,
0.0037909469101577997,
-0.016233881935477257,
-0.05179278552532196,
-0.061867620795965195,
-0.02382679656147957,
0.014660995453596115,
-0.06600694358348846,
0.0... |
120 | RAGTrack: Language-aware RGBT Tracking with Retrieval-Augmented Generation | [
"Hao Li",
"Yuhao Wang",
"Wenning Hao",
"Pingping Zhang",
"Dong Wang",
"Huchuan Lu"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Li_RAGTrack_Language-aware_RGBT_Tracking_with_Retrieval-Augmented_Generation_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Li_RAGTrack_Language-aware_RGBT_Tracking_with_Retrieval-Augmented_Generation_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Li_RAGTrack_Language-aware_RGBT_CVPR_2026_supplemental.pdf | 2603.03617 | cvf | @InProceedings{Li_2026_CVPR,
author = {Li, Hao and Wang, Yuhao and Hao, Wenning and Zhang, Pingping and Wang, Dong and Lu, Huchuan},
title = {RAGTrack: Language-aware RGBT Tracking with Retrieval-Augmented Generation},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern... | RGB-Thermal (RGBT) tracking aims to achieve robust object localization across diverse environmental conditions by fusing visible and thermal infrared modalities. However, existing RGBT trackers rely solely on initial-frame visual information for target modeling, failing to adapt to appearance variations due to the abse... | [
0.022026238963007927,
-0.023083897307515144,
0.013471262529492378,
0.05164889246225357,
0.0125601626932621,
0.005703990813344717,
0.020013175904750824,
0.028270503506064415,
-0.051105402410030365,
-0.037896398454904556,
-0.08926942944526672,
0.027116261422634125,
-0.06441093981266022,
-0.0... |
121 | Is the Modality Gap a Bug or a Feature? A Robustness Perspective | [
"Rhea Chowers",
"Oshri Naparstek",
"Udi Barzelay",
"Yair Weiss"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Chowers_Is_the_Modality_Gap_a_Bug_or_a_Feature_A_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Chowers_Is_the_Modality_Gap_a_Bug_or_a_Feature_A_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Chowers_Is_the_Modality_CVPR_2026_supplemental.pdf | 2603.29080 | cvf | @InProceedings{Chowers_2026_CVPR,
author = {Chowers, Rhea and Naparstek, Oshri and Barzelay, Udi and Weiss, Yair},
title = {Is the Modality Gap a Bug or a Feature? A Robustness Perspective},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
mo... | Many modern multi-modal models (e.g. CLIP) seek an embedding space in which the two modalities are aligned. Somewhat surprisingly, almost all existing models show a strong modality gap: the distribution of images is well-separated from the distribution of texts in the shared embedding space. Despite a series of recent ... | [
0.013921727426350117,
-0.015874402597546577,
0.016071874648332596,
0.06687071919441223,
0.03773113340139389,
-0.024147290736436844,
0.06793435662984848,
0.03294611722230911,
-0.0296011995524168,
-0.030824510380625725,
-0.03012016974389553,
0.014838830567896366,
-0.0736958459019661,
-0.0177... |
122 | Federated Active Learning Under Extreme Non-IID and Global Class Imbalance | [
"Chen-Chen Zong",
"Sheng-Jun Huang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Zong_Federated_Active_Learning_Under_Extreme_Non-IID_and_Global_Class_Imbalance_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Zong_Federated_Active_Learning_Under_Extreme_Non-IID_and_Global_Class_Imbalance_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zong_Federated_Active_Learning_CVPR_2026_supplemental.pdf | 2603.10341 | cvf | @InProceedings{Zong_2026_CVPR,
author = {Zong, Chen-Chen and Huang, Sheng-Jun},
title = {Federated Active Learning Under Extreme Non-IID and Global Class Imbalance},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
yea... | Federated active learning (FAL) seeks to reduce annotation cost under privacy constraints, yet its effectiveness degrades in realistic settings with severe global class imbalance and highly heterogeneous clients. We conduct a systematic study of query-model selection in FAL and uncover a central insight: the model that... | [
-0.001729674986563623,
-0.05908482149243355,
-0.005726574454456568,
0.013089973479509354,
0.00954938493669033,
-0.010946787893772125,
0.016177481040358543,
0.0018584263743832707,
-0.0245790034532547,
-0.03453761711716652,
-0.005235437303781509,
-0.009026787243783474,
-0.09362391382455826,
... |
123 | Tackling Alignment Ambiguity in Person Retrieval through Conversational Attribute Mining | [
"Hao Zou",
"Runqing Zhang",
"Jin Ding",
"Xue Zhou",
"Jianxiao Zou",
"Mingzhu Cai"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Zou_Tackling_Alignment_Ambiguity_in_Person_Retrieval_through_Conversational_Attribute_Mining_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Zou_Tackling_Alignment_Ambiguity_in_Person_Retrieval_through_Conversational_Attribute_Mining_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zou_Tackling_Alignment_Ambiguity_CVPR_2026_supplemental.pdf | null | null | @InProceedings{Zou_2026_CVPR,
author = {Zou, Hao and Zhang, Runqing and Ding, Jin and Zhou, Xue and Zou, Jianxiao and Cai, Mingzhu},
title = {Tackling Alignment Ambiguity in Person Retrieval through Conversational Attribute Mining},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Visi... | Text-to-Image Person Retrieval (TIPR) aims to retrieve pedestrian images with a given natural language description. It remains highly challenging due to the inherent ambiguity in cross-modal alignment: existing models often struggle to capture fine-grained correspondences, and their understanding of detailed pedestrian... | [
0.018052563071250916,
-0.033369068056344986,
-0.007693430874496698,
0.05766535550355911,
0.014306004159152508,
0.01643231511116028,
0.019913898780941963,
0.013348420150578022,
0.013891594484448433,
-0.04297473281621933,
-0.0656038224697113,
0.007332629524171352,
-0.09154118597507477,
-0.02... |
124 | Transform to Transfer: Boosting Adversarial Attack Transferability on Vision-Language Pre-training Models | [
"Yang Li",
"Jia-Li Yin",
"Luojun Lin",
"Wei Lin"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Li_Transform_to_Transfer_Boosting_Adversarial_Attack_Transferability_on_Vision-Language_Pre-training_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Li_Transform_to_Transfer_Boosting_Adversarial_Attack_Transferability_on_Vision-Language_Pre-training_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Li_Transform_to_Transfer_CVPR_2026_supplemental.pdf | null | null | @InProceedings{Li_2026_CVPR,
author = {Li, Yang and Yin, Jia-Li and Lin, Luojun and Lin, Wei},
title = {Transform to Transfer: Boosting Adversarial Attack Transferability on Vision-Language Pre-training Models},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogn... | Vision-Language Pre-training (VLP) models, while achieving state-of-the-art performance on various multimodal tasks, exhibit significant vulnerability to multimodal adversarial examples. In black-box attack scenarios of VLP models, a key challenge lies in the limited transferability of these adversarial examples. Exist... | [
-0.0084139509126544,
-0.02891559898853302,
-0.0014912016922608018,
0.0507890060544014,
0.034776922315359116,
0.020596487447619438,
0.03831649199128151,
0.0006362911663018167,
-0.009812761098146439,
-0.009812054224312305,
-0.03417232260107994,
0.004033438861370087,
-0.052711062133312225,
-0... |
125 | Streaming Diffusion Model for Fast Infrared and Visible Video Fusion | [
"Jinyuan Liu",
"Ludan Sun",
"Tengyu Ma",
"Chunyan Yang",
"Zhiying Jiang",
"Long Ma",
"Risheng Liu",
"Xin Fan"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Liu_Streaming_Diffusion_Model_for_Fast_Infrared_and_Visible_Video_Fusion_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Liu_Streaming_Diffusion_Model_for_Fast_Infrared_and_Visible_Video_Fusion_CVPR_2026_paper.pdf | null | null | null | @InProceedings{Liu_2026_CVPR,
author = {Liu, Jinyuan and Sun, Ludan and Ma, Tengyu and Yang, Chunyan and Jiang, Zhiying and Ma, Long and Liu, Risheng and Fan, Xin},
title = {Streaming Diffusion Model for Fast Infrared and Visible Video Fusion},
booktitle = {Proceedings of the IEEE/CVF Conference on C... | Infrared and visible video fusion is pivotal for robust perceptual systems, aiming to synthesize a comprehensive video stream that leverages both thermal resilience and textured details. However, prevailing methods, by treating videos as sequences of independent frames, inherently introduce temporal incoherence, such a... | [
0.015454797074198723,
0.007920720614492893,
-0.012665024027228355,
0.06757812201976776,
0.0389590710401535,
0.032624028623104095,
0.012620609253644943,
0.00496339937672019,
-0.01617659069597721,
-0.0865681916475296,
-0.01653469167649746,
-0.027736742049455643,
-0.04836256802082062,
0.01554... |
126 | Explore with Long-term Memory: A Benchmark and Multimodal LLM-based Reinforcement Learning Framework for Embodied Exploration | [
"Sen Wang",
"Bangwei Liu",
"Zhenkun Gao",
"Lizhuang Ma",
"Xuhong Wang",
"Yuan Xie",
"Xin Tan"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Wang_Explore_with_Long-term_Memory_A_Benchmark_and_Multimodal_LLM-based_Reinforcement_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Wang_Explore_with_Long-term_Memory_A_Benchmark_and_Multimodal_LLM-based_Reinforcement_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Wang_Explore_with_Long-term_CVPR_2026_supplemental.pdf | 2601.10744 | cvf | @InProceedings{Wang_2026_CVPR,
author = {Wang, Sen and Liu, Bangwei and Gao, Zhenkun and Ma, Lizhuang and Wang, Xuhong and Xie, Yuan and Tan, Xin},
title = {Explore with Long-term Memory: A Benchmark and Multimodal LLM-based Reinforcement Learning Framework for Embodied Exploration},
booktitle = {Pro... | An ideal embodied agent should possess lifelong learning capabilities to handle long-horizon and complex tasks, enabling continuous operation in general environments. This not only requires the agent to accurately accomplish given tasks but also to leverage long-term episodic memory to optimize decision-making. However... | [
-0.01599753275513649,
-0.005059356801211834,
-0.011851576156914234,
0.017466459423303604,
0.04646218195557594,
0.012295793741941452,
0.019433608278632164,
0.013944889418780804,
-0.04731006175279617,
-0.029310379177331924,
-0.02996058762073517,
0.01312674768269062,
-0.04178256168961525,
-0.... |
127 | Multi-Modal Image Fusion via Intervention-Stable Feature Learning | [
"Xue Wang",
"Zheng Guan",
"Wenhua Qian",
"Chengchao Wang",
"Runzhuo Ma"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Wang_Multi-Modal_Image_Fusion_via_Intervention-Stable_Feature_Learning_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Wang_Multi-Modal_Image_Fusion_via_Intervention-Stable_Feature_Learning_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Wang_Multi-Modal_Image_Fusion_CVPR_2026_supplemental.pdf | 2603.23272 | cvf | @InProceedings{Wang_2026_CVPR,
author = {Wang, Xue and Guan, Zheng and Qian, Wenhua and Wang, Chengchao and Ma, Runzhuo},
title = {Multi-Modal Image Fusion via Intervention-Stable Feature Learning},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}... | Multi-modal image fusion integrates complementary information from different modalities into a unified representation. Current methods predominantly optimize statistical correlations between modalities, often capturing dataset-induced spurious associations that degrade under distribution shifts. In this paper, we propo... | [
0.006030477117747068,
-0.026456499472260475,
0.0011476028012111783,
0.0413166843354702,
0.03784972056746483,
0.026717141270637512,
0.03157936781644821,
0.017672333866357803,
-0.051432009786367416,
-0.0631994754076004,
-0.020012712106108665,
0.006514702923595905,
-0.07998344302177429,
-0.01... |
128 | Dark3R: Learning Structure from Motion in the Dark | [
"Andrew Y. Guo",
"Anagh Malik",
"SaiKiran Tedla",
"Yutong Dai",
"Yiqian Qin",
"Zach Salehe",
"Benjamin Attal",
"Sotiris Nousias",
"Kiriakos N. Kutulakos",
"David B. Lindell"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Guo_Dark3R_Learning_Structure_from_Motion_in_the_Dark_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Guo_Dark3R_Learning_Structure_from_Motion_in_the_Dark_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Guo_Dark3R_Learning_Structure_CVPR_2026_supplemental.pdf | 2603.05330 | cvf | @InProceedings{Guo_2026_CVPR,
author = {Guo, Andrew Y. and Malik, Anagh and Tedla, SaiKiran and Dai, Yutong and Qin, Yiqian and Salehe, Zach and Attal, Benjamin and Nousias, Sotiris and Kutulakos, Kiriakos N. and Lindell, David B.},
title = {Dark3R: Learning Structure from Motion in the Dark},
bookti... | We introduce Dark3R, a framework for structure from motion in the dark that operates directly on raw images with signal-to-noise ratios (SNRs) below -4 dB--a regime where conventional feature- and learning-based methods break down. Our key insight is to adapt large-scale 3D foundation models to extreme low-light condit... | [
0.04622471332550049,
0.011483367532491684,
0.014757942408323288,
0.03627162054181099,
0.030270615592598915,
0.020049242302775383,
0.02116241119801998,
-0.000013895059055357706,
-0.06161624565720558,
-0.046861957758665085,
-0.006012722849845886,
0.004251738078892231,
-0.036660198122262955,
... |
129 | Gastric-X: A Multimodal Multi-Phase Benchmark Dataset for Advancing Vision-Language Models in Gastric Cancer Analysis | [
"Yuanzhe Li",
"Hao Chen",
"Rui Yin",
"Juyan Ba",
"Yu Zhang",
"Sheng Lu"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Li_Gastric-X_A_Multimodal_Multi-Phase_Benchmark_Dataset_for_Advancing_Vision-Language_Models_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Li_Gastric-X_A_Multimodal_Multi-Phase_Benchmark_Dataset_for_Advancing_Vision-Language_Models_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Li_Gastric-X_A_Multimodal_CVPR_2026_supplemental.pdf | 2603.19516 | cvf | @InProceedings{Li_2026_CVPR,
author = {Li, Yuanzhe and Chen, Hao and Yin, Rui and Ba, Juyan and Zhang, Yu and Lu, Sheng},
title = {Gastric-X: A Multimodal Multi-Phase Benchmark Dataset for Advancing Vision-Language Models in Gastric Cancer Analysis},
booktitle = {Proceedings of the IEEE/CVF Conferenc... | Recent vision-language models (VLMs) have shown strong generalization and multimodal reasoning abilities in natural domains. However, their application to medical diagnosis remains limited by the lack of comprehensive and structured datasets that capture real clinical workflows. To advance the development of VLMs for c... | [
-0.002978780772536993,
0.0194697268307209,
0.002734224544838071,
0.0020943409763276577,
0.05340860038995743,
-0.00014756442396901548,
0.009603713639080524,
0.025823671370744705,
-0.01673116907477379,
-0.05524659901857376,
-0.016934338957071304,
0.03714388608932495,
-0.048861850053071976,
-... |
130 | AVION: Aerial Vision-Language Instruction from Offline Teacher to Prompt-Tuned Network | [
"Yu Hu",
"Jianyang Gu",
"Hao Liu",
"Yue Cao",
"Jozsef Hamari",
"Zheng Liu",
"Mohsen Zardadi"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Hu_AVION_Aerial_Vision-Language_Instruction_from_Offline_Teacher_to_Prompt-Tuned_Network_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Hu_AVION_Aerial_Vision-Language_Instruction_from_Offline_Teacher_to_Prompt-Tuned_Network_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Hu_AVION_Aerial_Vision-Language_CVPR_2026_supplemental.pdf | 2603.12659 | cvf | @InProceedings{Hu_2026_CVPR,
author = {Hu, Yu and Gu, Jianyang and Liu, Hao and Cao, Yue and Hamari, Jozsef and Liu, Zheng and Zardadi, Mohsen},
title = {AVION: Aerial Vision-Language Instruction from Offline Teacher to Prompt-Tuned Network},
booktitle = {Proceedings of the IEEE/CVF Conference on Com... | Adapting vision-language models to remote sensing imagery remains challenging due to two key factors: limited semantic coverage in textual representations and insufficient adaptability of visual features. These issues are particularly significant in aerial scenes, which involve various visual appearances and fine-grain... | [
0.029144233092665672,
-0.00182166020385921,
0.0021635964512825012,
0.0496099591255188,
0.017347512766718864,
-0.007514351513236761,
0.044924862682819366,
0.00953513104468584,
-0.028182078152894974,
-0.005818115547299385,
-0.06626071780920029,
0.01568830944597721,
-0.08916565775871277,
-0.0... |
131 | Efficient and Training-Free Single-Image Diffusion Models | [
"Haojun Qiu",
"Kiriakos N. Kutulakos",
"David B. Lindell"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Qiu_Efficient_and_Training-Free_Single-Image_Diffusion_Models_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Qiu_Efficient_and_Training-Free_Single-Image_Diffusion_Models_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Qiu_Efficient_and_Training-Free_CVPR_2026_supplemental.pdf | null | null | @InProceedings{Qiu_2026_CVPR,
author = {Qiu, Haojun and Kutulakos, Kiriakos N. and Lindell, David B.},
title = {Efficient and Training-Free Single-Image Diffusion Models},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
... | We consider the problem of generating images whose internal structure - defined by the distribution of patches across multiple scales---matches that of a single reference image. Recent approaches address this problem by training a diffusion model on a single image. But even in this setting, training is computationally ... | [
0.0036547447089105844,
-0.009134793654084206,
0.008986104279756546,
0.055314064025878906,
0.05510841682553291,
0.043256841599941254,
0.005951466038823128,
-0.0011189525248482823,
-0.019735733047127724,
-0.07852914184331894,
-0.00004340857412898913,
-0.004682796075940132,
-0.05009708181023598... |
132 | OSA: Echocardiography Video Segmentation via Orthogonalized State Update and Anatomical Prior-aware Feature Enhancement | [
"Rui Wang",
"Huisi Wu",
"Jing Qin"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Wang_OSA_Echocardiography_Video_Segmentation_via_Orthogonalized_State_Update_and_Anatomical_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Wang_OSA_Echocardiography_Video_Segmentation_via_Orthogonalized_State_Update_and_Anatomical_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Wang_OSA_Echocardiography_Video_CVPR_2026_supplemental.pdf | 2603.26188 | cvf | @InProceedings{Wang_2026_CVPR,
author = {Wang, Rui and Wu, Huisi and Qin, Jing},
title = {OSA: Echocardiography Video Segmentation via Orthogonalized State Update and Anatomical Prior-aware Feature Enhancement},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogn... | Accurate and temporally consistent segmentation of the left ventricle from echocardiography videos is essential for estimating the ejection fraction and assessing cardiac function. However, modeling spatiotemporal dynamics remains difficult due to severe speckle noise and rapid non-rigid deformations. Existing linear r... | [
-0.005564770195633173,
-0.009273996576666832,
0.0024086209014058113,
0.00596240209415555,
0.04660326614975929,
0.06187200918793678,
0.029493551701307297,
0.0423259437084198,
-0.05857400223612785,
-0.05303937941789627,
0.012863481417298317,
-0.0313236340880394,
-0.05196307972073555,
-0.0135... |
133 | Detecting Compressed AI-Generated Images via Phase Spectrum Robustness | [
"Kai Li",
"Wenqi Ren",
"Wei Wang",
"Xiaochun Cao"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Li_Detecting_Compressed_AI-Generated_Images_via_Phase_Spectrum_Robustness_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Li_Detecting_Compressed_AI-Generated_Images_via_Phase_Spectrum_Robustness_CVPR_2026_paper.pdf | null | null | null | @InProceedings{Li_2026_CVPR,
author = {Li, Kai and Ren, Wenqi and Wang, Wei and Cao, Xiaochun},
title = {Detecting Compressed AI-Generated Images via Phase Spectrum Robustness},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {Ju... | This paper aims to present a robust AI-generated image detection framework designed to address performance degradation caused by image compression in online social networks. The key challenges are twofold: 1) compression destroys fragile artifacts that are crucial to existing methods, and 2) it introduces new compressi... | [
0.01763647235929966,
-0.041787032037973404,
0.006496300455182791,
0.025518950074911118,
0.042591772973537445,
0.017075343057513237,
0.015770960599184036,
-0.008827242068946362,
-0.038644611835479736,
-0.06794819235801697,
-0.030967343598604202,
-0.01771237887442112,
-0.04638925939798355,
-... |
134 | VLM-Guided Group Preference Alignment for Diffusion-based Human Mesh Recovery | [
"Wenhao Shen",
"Hao Wang",
"Wanqi Yin",
"Fayao Liu",
"Xulei Yang",
"Chao Liang",
"Zhongang Cai",
"Guosheng Lin"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Shen_VLM-Guided_Group_Preference_Alignment_for_Diffusion-based_Human_Mesh_Recovery_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Shen_VLM-Guided_Group_Preference_Alignment_for_Diffusion-based_Human_Mesh_Recovery_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Shen_VLM-Guided_Group_Preference_CVPR_2026_supplemental.pdf | 2602.19180 | cvf | @InProceedings{Shen_2026_CVPR,
author = {Shen, Wenhao and Wang, Hao and Yin, Wanqi and Liu, Fayao and Yang, Xulei and Liang, Chao and Cai, Zhongang and Lin, Guosheng},
title = {VLM-Guided Group Preference Alignment for Diffusion-based Human Mesh Recovery},
booktitle = {Proceedings of the IEEE/CVF Con... | Human mesh recovery (HMR) from a single RGB image is inherently ambiguous, as multiple 3D poses can correspond to the same 2D observation. Recent diffusion-based methods tackle this by generating various hypotheses, but often sacrifice accuracy. They yield predictions that are either physically implausible or drift fro... | [
0.026872893795371056,
0.03525150939822197,
0.022052645683288574,
0.036143381148576736,
0.026349041610956192,
0.014774407260119915,
0.027036869898438454,
-0.004539116285741329,
-0.028003649786114693,
-0.09290466457605362,
-0.015576778911054134,
-0.007025355007499456,
-0.038711100816726685,
... |
135 | Tunable Soft Equivariance with Guarantees | [
"Md Ashiqur Rahman",
"Lim Jun Hao",
"Jeremiah Jiang",
"Teck-Yian Lim",
"Raymond A. Yeh"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Rahman_Tunable_Soft_Equivariance_with_Guarantees_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Rahman_Tunable_Soft_Equivariance_with_Guarantees_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Rahman_Tunable_Soft_Equivariance_CVPR_2026_supplemental.pdf | 2603.26657 | cvf | @InProceedings{Rahman_2026_CVPR,
author = {Rahman, Md Ashiqur and Hao, Lim Jun and Jiang, Jeremiah and Lim, Teck-Yian and Yeh, Raymond A.},
title = {Tunable Soft Equivariance with Guarantees},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
... | Equivariance is a fundamental property in computer vision models, yet strict equivariance is rarely satisfied in real-world data, which can limit a model's performance. Controlling the degree of equivariance is therefore desirable. We propose a general framework for constructing soft equivariant models by projecting th... | [
0.029033198952674866,
-0.017566848546266556,
0.027871517464518547,
0.013770662248134613,
0.02295711077749729,
0.03285288065671921,
0.01941138319671154,
0.027262268587946892,
-0.038049980998039246,
-0.041275959461927414,
-0.04604995623230934,
-0.017334992066025734,
-0.08178084343671799,
0.0... |
136 | PhysIR-Splat: Physically Consistent Thermal Infrared Radiative Transfer in 3D Gaussian Splatting | [
"Jingyuan Gao",
"Yumeng Hu",
"Fei Gao",
"Mingjin Zhang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Gao_PhysIR-Splat_Physically_Consistent_Thermal_Infrared_Radiative_Transfer_in_3D_Gaussian_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Gao_PhysIR-Splat_Physically_Consistent_Thermal_Infrared_Radiative_Transfer_in_3D_Gaussian_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Gao_PhysIR-Splat_Physically_Consistent_CVPR_2026_supplemental.pdf | null | null | @InProceedings{Gao_2026_CVPR,
author = {Gao, Jingyuan and Hu, Yumeng and Gao, Fei and Zhang, Mingjin},
title = {PhysIR-Splat: Physically Consistent Thermal Infrared Radiative Transfer in 3D Gaussian Splatting},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni... | Thermal infrared (TIR) 3D reconstruction provides geometry that is intrinsically coupled to the temperature field, even in low-light, nighttime, and smoke-obscured environments. TIR imaging measures self-emitted thermal radiation driven by object temperature and is largely independent of external illumination; therefor... | [
0.031657151877880096,
0.027565719559788704,
0.014169539324939251,
0.02701592445373535,
0.021825209259986877,
-0.00598099734634161,
0.008378688246011734,
-0.009392554871737957,
-0.017895402386784554,
-0.05257278308272362,
-0.03839442878961563,
-0.002791572595015168,
-0.03971772640943527,
0.... |
137 | Generalized-CVO: Fast and Correspondence-Free Local Point Cloud Registration with Second Order Riemannian Optimization | [
"Ray Zhang",
"Marcus Greiff",
"Thomas Lew",
"John Subosits"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Zhang_Generalized-CVO_Fast_and_Correspondence-Free_Local_Point_Cloud_Registration_with_Second_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Zhang_Generalized-CVO_Fast_and_Correspondence-Free_Local_Point_Cloud_Registration_with_Second_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zhang_Generalized-CVO_Fast_and_CVPR_2026_supplemental.pdf | null | null | @InProceedings{Zhang_2026_CVPR,
author = {Zhang, Ray and Greiff, Marcus and Lew, Thomas and Subosits, John},
title = {Generalized-CVO: Fast and Correspondence-Free Local Point Cloud Registration with Second Order Riemannian Optimization},
booktitle = {Proceedings of the IEEE/CVF Conference on Compute... | We propose a fast and correspondence-free local point cloud registration method that leverages geometric surface structure and reproducing kernel Hilbert space (RKHS) embeddings. The method represents point clouds as continuous functions with point-wise anisotropic kernels that encode local geometry. This formulation i... | [
0.006049217656254768,
0.018752414733171463,
0.056258466094732285,
0.031659916043281555,
0.018469462171196938,
0.07157464325428009,
0.013011748902499676,
0.027866052463650703,
-0.02267378382384777,
-0.05459306016564369,
-0.046659521758556366,
-0.005374970845878124,
-0.06454873085021973,
0.0... |
138 | MultiAnimate: Pose-Guided Image Animation Made Extensible | [
"Yingcheng Hu",
"Haowen Gong",
"Chuanguang Yang",
"Zhulin An",
"Yongjun Xu",
"Songhua Liu"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Hu_MultiAnimate_Pose-Guided_Image_Animation_Made_Extensible_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Hu_MultiAnimate_Pose-Guided_Image_Animation_Made_Extensible_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Hu_MultiAnimate_Pose-Guided_Image_CVPR_2026_supplemental.pdf | 2602.21581 | cvf | @InProceedings{Hu_2026_CVPR,
author = {Hu, Yingcheng and Gong, Haowen and Yang, Chuanguang and An, Zhulin and Xu, Yongjun and Liu, Songhua},
title = {MultiAnimate: Pose-Guided Image Animation Made Extensible},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognit... | Pose-guided human image animation aims to synthesize realistic videos of a reference character driven by a sequence of poses. While diffusion-based methods have achieved remarkable success, most existing approaches are limited to single-character animation. We observe that naively extending these methods to multi-chara... | [
-0.0028885225765407085,
-0.0322384312748909,
-0.025069957599043846,
0.04229062423110008,
0.03096616081893444,
0.03138384968042374,
0.004697486292570829,
-0.001089997822418809,
-0.023075802251696587,
-0.05712982267141342,
-0.037613287568092346,
-0.01983192190527916,
-0.03953346982598305,
-0... |
139 | Exposing and Evaluating Hallucinations for GUI Grounding | [
"Zicheng Zhang",
"Hongyi Jing",
"Rui Lv",
"Shuo Fang",
"Shiai Zhu",
"Junying Wang",
"Chunyi Li",
"Xiaohong Liu",
"Chenguang Ma",
"Guangtao Zhai"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Zhang_Exposing_and_Evaluating_Hallucinations_for_GUI_Grounding_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Zhang_Exposing_and_Evaluating_Hallucinations_for_GUI_Grounding_CVPR_2026_paper.pdf | null | null | null | @InProceedings{Zhang_2026_CVPR,
author = {Zhang, Zicheng and Jing, Hongyi and Lv, Rui and Fang, Shuo and Zhu, Shiai and Wang, Junying and Li, Chunyi and Liu, Xiaohong and Ma, Chenguang and Zhai, Guangtao},
title = {Exposing and Evaluating Hallucinations for GUI Grounding},
booktitle = {Proceedings of... | Existing GUI benchmarks primarily focus on evaluating models' comprehensive capabilities but largely overlook hallucination phenomena in grounding tasks, which are crucial to the reliability of GUI understanding. In this work, we expose two major types of hallucinations in GUI grounding: 1) Confusion Hallucination, whe... | [
-0.001438062870875001,
0.03190426528453827,
-0.030115123838186264,
0.019535232335329056,
0.004123562015593052,
-0.0007881465135142207,
0.04424082860350609,
0.025699226185679436,
-0.03326123580336571,
-0.046904049813747406,
-0.040795352309942245,
0.020570408552885056,
-0.08483820408582687,
... |
140 | Beyond Sequential Tools: A Unified VLM Agent System for Photographic Post-Processing via Dynamic Multi-Expert Fusion | [
"Honglin Xiong",
"Chenjie Zhu",
"Jianbiao Ding",
"Zixuan Ni",
"Wei Li",
"Zhenpeng Mi",
"Qian Wang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Xiong_Beyond_Sequential_Tools_A_Unified_VLM_Agent_System_for_Photographic_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Xiong_Beyond_Sequential_Tools_A_Unified_VLM_Agent_System_for_Photographic_CVPR_2026_paper.pdf | null | null | null | @InProceedings{Xiong_2026_CVPR,
author = {Xiong, Honglin and Zhu, Chenjie and Ding, Jianbiao and Ni, Zixuan and Li, Wei and Mi, Zhenpeng and Wang, Qian},
title = {Beyond Sequential Tools: A Unified VLM Agent System for Photographic Post-Processing via Dynamic Multi-Expert Fusion},
booktitle = {Procee... | Real-world image restoration is challenged by complex, coupled degradations. Existing "all-in-one" models often lack generalization, while agentic systems suffer from inefficient sequential tool invocation. We propose a VLM-guided one-shot framework for universal photographic post-processing. Our system employs a Visio... | [
0.04119989648461342,
-0.017833072692155838,
0.018577037379145622,
0.05365147441625595,
0.0505911149084568,
0.01650300994515419,
0.03791338950395584,
-0.00940578430891037,
-0.021026387810707092,
-0.05768755078315735,
-0.03128916770219803,
0.03424501419067383,
-0.06488129496574402,
-0.017629... |
141 | ReAttnCLIP: Training-Free Open-Vocabulary Remote Sensing Image Segmentation via Re-defined Attention in CLIP | [
"Xin Niu",
"Manqi Zhao",
"Dongsheng Jiang",
"Yingying Wu",
"Bing Su"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Niu_ReAttnCLIP_Training-Free_Open-Vocabulary_Remote_Sensing_Image_Segmentation_via_Re-defined_Attention_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Niu_ReAttnCLIP_Training-Free_Open-Vocabulary_Remote_Sensing_Image_Segmentation_via_Re-defined_Attention_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Niu_ReAttnCLIP_Training-Free_Open-Vocabulary_CVPR_2026_supplemental.pdf | null | null | @InProceedings{Niu_2026_CVPR,
author = {Niu, Xin and Zhao, Manqi and Jiang, Dongsheng and Wu, Yingying and Su, Bing},
title = {ReAttnCLIP: Training-Free Open-Vocabulary Remote Sensing Image Segmentation via Re-defined Attention in CLIP},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer... | Remote sensing image segmentation is critical for a range of applications, including natural disaster monitoring and precision agriculture. Open-vocabulary segmentation enhances flexibility by removing fixed category constraints, enabling more fine-grained and adaptive scene understanding. Unlike CLIP's original pretra... | [
0.02171684429049492,
-0.042186155915260315,
0.020989684388041496,
0.044828105717897415,
0.017801953479647636,
0.028128929436206818,
0.035027772188186646,
0.025984670966863632,
-0.014294521883130074,
-0.03608985245227814,
-0.0769384503364563,
0.02626059018075466,
-0.08472148329019547,
-0.01... |
142 | LIBERO-Plus: A Progressive Robustness Benchmark for Visual-Language-Action Models | [
"Senyu Fei",
"Siyin Wang",
"Junhao Shi",
"Zihao Dai",
"Jikun Cai",
"Pengfang Qian",
"Li Ji",
"Xinzhe He",
"Shiduo Zhang",
"Zhaoye Fei",
"Jinlan Fu",
"Jingjing Gong",
"Xipeng Qiu"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Fei_LIBERO-Plus_A_Progressive_Robustness_Benchmark_for_Visual-Language-Action_Models_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Fei_LIBERO-Plus_A_Progressive_Robustness_Benchmark_for_Visual-Language-Action_Models_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Fei_LIBERO-Plus_A_Progressive_CVPR_2026_supplemental.pdf | 2510.13626 | title_judge | @InProceedings{Fei_2026_CVPR,
author = {Fei, Senyu and Wang, Siyin and Shi, Junhao and Dai, Zihao and Cai, Jikun and Qian, Pengfang and Ji, Li and He, Xinzhe and Zhang, Shiduo and Fei, Zhaoye and Fu, Jinlan and Gong, Jingjing and Qiu, Xipeng},
title = {LIBERO-Plus: A Progressive Robustness Benchmark for ... | Visual-Language-Action (VLA) models report impressive success rates exceeding 95% on robotic manipulation benchmarks, yet these results may mask fundamental weaknesses in robustness. Current simulation-based robustness evaluations suffer from narrow perturbation coverage, manual design constraints, and coarse-grained a... | [
0.01624247059226036,
0.010846040211617947,
0.010112602263689041,
0.0257892869412899,
0.05107351392507553,
0.027133625000715256,
0.028267187997698784,
0.001423862879164517,
-0.03744035214185715,
-0.041475869715213776,
-0.00480053573846817,
0.0006643596570938826,
-0.05766341835260391,
-0.002... |
143 | MajutsuCity: Language-driven Aesthetic-adaptive City Generation with Controllable 3D Assets and Layouts | [
"Zilong Huang",
"Jun He",
"Xiaobin Huang",
"Ziyi Xiong",
"Yang Luo",
"Junyan Ye",
"Weijia Li",
"Yiping Chen",
"Ting Han"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Huang_MajutsuCity_Language-driven_Aesthetic-adaptive_City_Generation_with_Controllable_3D_Assets_and_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Huang_MajutsuCity_Language-driven_Aesthetic-adaptive_City_Generation_with_Controllable_3D_Assets_and_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Huang_MajutsuCity_Language-driven_Aesthetic-adaptive_CVPR_2026_supplemental.pdf | 2511.20415 | cvf | @InProceedings{Huang_2026_CVPR,
author = {Huang, Zilong and He, Jun and Huang, Xiaobin and Xiong, Ziyi and Luo, Yang and Ye, Junyan and Li, Weijia and Chen, Yiping and Han, Ting},
title = {MajutsuCity: Language-driven Aesthetic-adaptive City Generation with Controllable 3D Assets and Layouts},
bookti... | Generating realistic 3D cities is fundamental to world models, virtual reality, and game development, where an ideal urban scene must satisfy both stylistic diversity, fine-grained, and controllability. However, existing methods struggle to balance the creative flexibility offered by text-based generation with the obje... | [
-0.02168886363506317,
-0.04645797610282898,
0.02532578632235527,
0.034580208361148834,
0.027763918042182922,
0.04284903034567833,
0.03936561197042465,
0.04115848243236542,
-0.015271738171577454,
-0.056387875229120255,
-0.06491672992706299,
-0.022912193089723587,
-0.08414636552333832,
0.008... |
144 | AnyLift: Scaling Motion Reconstruction from Internet Videos via 2D Diffusion | [
"Hongjie Li",
"Heng Yu",
"Jiaman Li",
"Hong-Xing Yu",
"Ehsan Adeli",
"C. Karen Liu",
"Jiajun Wu"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Li_AnyLift_Scaling_Motion_Reconstruction_from_Internet_Videos_via_2D_Diffusion_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Li_AnyLift_Scaling_Motion_Reconstruction_from_Internet_Videos_via_2D_Diffusion_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Li_AnyLift_Scaling_Motion_CVPR_2026_supplemental.zip | 2604.17818 | cvf | @InProceedings{Li_2026_CVPR,
author = {Li, Hongjie and Yu, Heng and Li, Jiaman and Yu, Hong-Xing and Adeli, Ehsan and Liu, C. Karen and Wu, Jiajun},
title = {AnyLift: Scaling Motion Reconstruction from Internet Videos via 2D Diffusion},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer ... | Reconstructing 3D human motion and human-object interactions (HOI) from Internet videos is a fundamental step toward building large-scale datasets of human behavior. Existing methods struggle to recover globally consistent 3D motion under dynamic cameras, especially for motion types underrepresented in current motion-c... | [
-0.010155095718801022,
-0.0004094445030204952,
-0.004077852237969637,
0.023242685943841934,
0.04487311840057373,
0.021377773955464363,
0.034295372664928436,
0.009333627298474312,
-0.03583521395921707,
-0.043168485164642334,
-0.008943520486354828,
-0.06121325492858887,
-0.06563249230384827,
... |
145 | Accelerating Diffusion via Hybrid Data-Pipeline Parallelism Based on Conditional Guidance Scheduling | [
"Euisoo Jung",
"Byunghyun Kim",
"Hyunjin Kim",
"Seonghye Cho",
"Jae-Gil Lee"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Jung_Accelerating_Diffusion_via_Hybrid_Data-Pipeline_Parallelism_Based_on_Conditional_Guidance_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Jung_Accelerating_Diffusion_via_Hybrid_Data-Pipeline_Parallelism_Based_on_Conditional_Guidance_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Jung_Accelerating_Diffusion_via_CVPR_2026_supplemental.pdf | 2602.21760 | cvf | @InProceedings{Jung_2026_CVPR,
author = {Jung, Euisoo and Kim, Byunghyun and Kim, Hyunjin and Cho, Seonghye and Lee, Jae-Gil},
title = {Accelerating Diffusion via Hybrid Data-Pipeline Parallelism Based on Conditional Guidance Scheduling},
booktitle = {Proceedings of the IEEE/CVF Conference on Compute... | Diffusion models have achieved remarkable progress in high-fidelity image, video, and audio generation, yet inference remains computationally expensive. Nevertheless, current diffusion acceleration methods based on distributed parallelism suffer from noticeable generation artifacts and fail to achieve substantial accel... | [
-0.024122528731822968,
-0.02924293465912342,
-0.0018295769114047289,
0.06961337476968765,
0.030555076897144318,
0.07401300221681595,
0.0029598490800708532,
-0.005882204044610262,
-0.003809574292972684,
-0.05962256342172623,
0.008490275591611862,
-0.02353937178850174,
-0.025171972811222076,
... |
146 | OpenDance: Multimodal Controllable 3D Dance Generation with Large-scale Internet Data | [
"Jinlu Zhang",
"Zixi Kang",
"Libin Liu",
"Jianlong Chang",
"Qi Tian",
"Feng Gao",
"Yizhou Wang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Zhang_OpenDance_Multimodal_Controllable_3D_Dance_Generation_with_Large-scale_Internet_Data_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Zhang_OpenDance_Multimodal_Controllable_3D_Dance_Generation_with_Large-scale_Internet_Data_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zhang_OpenDance_Multimodal_Controllable_CVPR_2026_supplemental.zip | 2506.07565 | cvf | @InProceedings{Zhang_2026_CVPR,
author = {Zhang, Jinlu and Kang, Zixi and Liu, Libin and Chang, Jianlong and Tian, Qi and Gao, Feng and Wang, Yizhou},
title = {OpenDance: Multimodal Controllable 3D Dance Generation with Large-scale Internet Data},
booktitle = {Proceedings of the IEEE/CVF Conference o... | Music-driven 3D dance generation offers significant creative potential, yet practical applications demand versatile and multimodal control. Given the highly dynamic and complex human motion covering various styles and genres, dance generation requires satisfying diverse conditions beyond just music (e.g., spatial traje... | [
-0.011099928058683872,
-0.021152470260858536,
-0.0036408358719199896,
0.07361963391304016,
0.03325594216585159,
-0.016755083575844765,
0.04495742544531822,
0.01436708029359579,
-0.04636969044804573,
-0.05011947080492973,
-0.043914712965488434,
-0.029998980462551117,
-0.07711712270975113,
-... |
147 | UniVerse: Empower Unified Generation with Reasoning and Knowledge | [
"Kaiyue Sun",
"Weiyang Jin",
"Chengqi Duan",
"Rongyao Fang",
"Xian Liu",
"Yuwei Niu",
"Chunwei Wang",
"Aoxue Li",
"Xihui Liu"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Sun_UniVerse_Empower_Unified_Generation_with_Reasoning_and_Knowledge_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Sun_UniVerse_Empower_Unified_Generation_with_Reasoning_and_Knowledge_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Sun_UniVerse_Empower_Unified_CVPR_2026_supplemental.pdf | null | null | @InProceedings{Sun_2026_CVPR,
author = {Sun, Kaiyue and Jin, Weiyang and Duan, Chengqi and Fang, Rongyao and Liu, Xian and Niu, Yuwei and Wang, Chunwei and Li, Aoxue and Liu, Xihui},
title = {UniVerse: Empower Unified Generation with Reasoning and Knowledge},
booktitle = {Proceedings of the IEEE/CVF ... | Current text-to-image (T2I) generation models often struggle with prompts that require complex reasoning or specialized knowledge, failing to accurately interpret implicit user intent. To bridge this gap, we introduce T2I-Reason, a large-scale dataset designed to empower text-to-image generation in unified multimodal m... | [
0.0057944031432271,
-0.029142096638679504,
0.004208772908896208,
0.06545159220695496,
0.04489404708147049,
0.005544676445424557,
0.026067683473229408,
0.015617290511727333,
-0.02545679733157158,
-0.001674432191066444,
-0.04906878247857094,
0.03650903329253197,
-0.07072553783655167,
-0.0105... |
148 | EcoAlign: An Economically Rational Framework for Efficient LVLM Alignment | [
"Ruoxi Cheng",
"Hao-Xuan Ma",
"Teng Ma",
"Hongyi Zhang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Cheng_EcoAlign_An_Economically_Rational_Framework_for_Efficient_LVLM_Alignment_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Cheng_EcoAlign_An_Economically_Rational_Framework_for_Efficient_LVLM_Alignment_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Cheng_EcoAlign_An_Economically_CVPR_2026_supplemental.pdf | 2511.11301 | cvf | @InProceedings{Cheng_2026_CVPR,
author = {Cheng, Ruoxi and Ma, Hao-Xuan and Ma, Teng and Zhang, Hongyi},
title = {EcoAlign: An Economically Rational Framework for Efficient LVLM Alignment},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
mon... | Large Vision-Language Models (LVLMs) exhibit powerful reasoning capabilities but suffer sophisticated jailbreak vulnerabilities. Fundamentally, aligning LVLMs is not just a safety challenge but a problem of economic efficiency. Current alignment methods struggle with the trade-off between safety, utility, and operation... | [
0.014955630525946617,
0.003373374231159687,
-0.006728359963744879,
0.035066984593868256,
0.027143046259880066,
0.03944156691431999,
0.03174128010869026,
0.002156623173505068,
-0.02229946292936802,
-0.02344471588730812,
-0.01471723522990942,
0.03611768037080765,
-0.09346676617860794,
-0.032... |
149 | Inside-Out: Measuring Generalization in Vision Transformers Through Inner Workings | [
"Yunxiang Peng",
"Mengmeng Ma",
"Ziyu Yao",
"Xi Peng"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Peng_Inside-Out_Measuring_Generalization_in_Vision_Transformers_Through_Inner_Workings_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Peng_Inside-Out_Measuring_Generalization_in_Vision_Transformers_Through_Inner_Workings_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Peng_Inside-Out_Measuring_Generalization_CVPR_2026_supplemental.pdf | 2604.08192 | cvf | @InProceedings{Peng_2026_CVPR,
author = {Peng, Yunxiang and Ma, Mengmeng and Yao, Ziyu and Peng, Xi},
title = {Inside-Out: Measuring Generalization in Vision Transformers Through Inner Workings},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
... | Reliable generalization metrics are fundamental to the evaluation of machine learning models. Especially in high-stakes applications where labeled target data are scarce, evaluation of models' generalization performance under distribution shift is a pressing need. We focus on two practical scenarios: (1) Before deploym... | [
0.012894484214484692,
-0.04381759092211723,
-0.017672909423708916,
0.025903245434165,
0.05197124555706978,
0.006454773247241974,
0.0368167869746685,
-0.01248984970152378,
-0.012553599663078785,
-0.02043836936354637,
-0.01647385209798813,
0.020596904680132866,
-0.05545695871114731,
-0.00743... |
150 | AniMimic: Imitating 3D Animation from Video Priors | [
"Tianyi Xie",
"Yunuo Chen",
"Yaowei Guo",
"Yin Yang",
"Bolei Zhou",
"Demetri Terzopoulos",
"Ying Jiang",
"Chenfanfu Jiang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Xie_AniMimic_Imitating_3D_Animation_from_Video_Priors_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Xie_AniMimic_Imitating_3D_Animation_from_Video_Priors_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Xie_AniMimic_Imitating_3D_CVPR_2026_supplemental.pdf | 2512.14133 | title_judge | @InProceedings{Xie_2026_CVPR,
author = {Xie, Tianyi and Chen, Yunuo and Guo, Yaowei and Yang, Yin and Zhou, Bolei and Terzopoulos, Demetri and Jiang, Ying and Jiang, Chenfanfu},
title = {AniMimic: Imitating 3D Animation from Video Priors},
booktitle = {Proceedings of the IEEE/CVF Conference on Comput... | Creating realistic 3D animation remains a time-consuming and expertise-dependent process, requiring manual rigging, keyframing, and fine-tuning of complex motions. Meanwhile, video diffusion models have recently demonstrated remarkable 2D motion imagination, generating dynamic and visually coherent motion from text or ... | [
0.016958732157945633,
-0.0014338430482894182,
-0.013593219220638275,
0.029989540576934814,
0.04811325669288635,
0.044169746339321136,
0.016805626451969147,
-0.02358655259013176,
-0.03681964427232742,
-0.08819182217121124,
0.0021825802978128195,
-0.031052744016051292,
-0.012554997578263283,
... |
151 | Building Robust Vision Encoders for Cross-Dataset Evaluation in Immunofluorescent Microscopy | [
"Umar Marikkar",
"Syed Sameed Husain",
"Muhammad Awais",
"Sara Atito"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Marikkar_Building_Robust_Vision_Encoders_for_Cross-Dataset_Evaluation_in_Immunofluorescent_Microscopy_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Marikkar_Building_Robust_Vision_Encoders_for_Cross-Dataset_Evaluation_in_Immunofluorescent_Microscopy_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Marikkar_Building_Robust_Vision_CVPR_2026_supplemental.pdf | null | null | @InProceedings{Marikkar_2026_CVPR,
author = {Marikkar, Umar and Husain, Syed Sameed and Awais, Muhammad and Atito, Sara},
title = {Building Robust Vision Encoders for Cross-Dataset Evaluation in Immunofluorescent Microscopy},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and ... | Immunofluorescence (IF) images reveal detailed information about structures and functions at the subcellular level. However, unlike RGB images, IF datasets pose challenges for deep learning models due to their inconsistencies in channel count and configuration, stemming from varying staining protocols across laboratori... | [
0.010586639866232872,
-0.0024005211889743805,
-0.008471199311316013,
0.03054075501859188,
0.039195772260427475,
0.005977978464215994,
0.010661102831363678,
-0.010688538663089275,
-0.016051774844527245,
-0.04339561611413956,
-0.003327746409922838,
-0.019538871943950653,
-0.06766519695520401,
... |
152 | Gated KalmaNet: A Fading Memory Layer through Test-time Ridge Regression | [
"Liangzu Peng",
"Aditya Chattopadhyay",
"Luca Zancato",
"Elvis Nunez",
"Wei Xia",
"Stefano Soatto"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Peng_Gated_KalmaNet_A_Fading_Memory_Layer_through_Test-time_Ridge_Regression_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Peng_Gated_KalmaNet_A_Fading_Memory_Layer_through_Test-time_Ridge_Regression_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Peng_Gated_KalmaNet_A_CVPR_2026_supplemental.pdf | 2511.21016 | cvf | @InProceedings{Peng_2026_CVPR,
author = {Peng, Liangzu and Chattopadhyay, Aditya and Zancato, Luca and Nunez, Elvis and Xia, Wei and Soatto, Stefano},
title = {Gated KalmaNet: A Fading Memory Layer through Test-time Ridge Regression},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vi... | As efficient alternatives to softmax Attention, linear state space models (SSMs) achieve constant memory and linear compute, but maintain only a lossy, fading summary of the past, often leading to inferior performance in recall oriented settings. We propose Gated KalmaNet (GKA), a layer that reduces this gap by account... | [
-0.04163070023059845,
-0.03778815641999245,
0.01364077813923359,
0.036257773637771606,
0.03663073480129242,
0.032426781952381134,
0.038098908960819244,
0.0103891771286726,
-0.009926119819283485,
-0.02459055557847023,
0.008610589429736137,
0.007023324258625507,
-0.06494428217411041,
0.00197... |
153 | The Missing Point in Vision Transformers for Universal Image Segmentation | [
"Sajjad Shahabodini",
"Mobina Mansoori",
"Farnoush Bayatmakou",
"Jamshid Abouei",
"Konstantinos Plataniotis",
"Arash Mohammadi"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Shahabodini_The_Missing_Point_in_Vision_Transformers_for_Universal_Image_Segmentation_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Shahabodini_The_Missing_Point_in_Vision_Transformers_for_Universal_Image_Segmentation_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Shahabodini_The_Missing_Point_CVPR_2026_supplemental.zip | 2505.19795 | cvf | @InProceedings{Shahabodini_2026_CVPR,
author = {Shahabodini, Sajjad and Mansoori, Mobina and Bayatmakou, Farnoush and Abouei, Jamshid and Plataniotis, Konstantinos and Mohammadi, Arash},
title = {The Missing Point in Vision Transformers for Universal Image Segmentation},
booktitle = {Proceedings of t... | Image segmentation remains a challenging task in computer vision, demanding robust mask generation and precise classification. Recent mask-based approaches yield high-quality masks by capturing global context. However, accurately classifying these masks, especially in the presence of ambiguous boundaries and imbalanced... | [
-0.01429068110883236,
-0.006405627354979515,
0.008410749956965446,
0.025908851996064186,
0.013278008438646793,
0.06352365016937256,
0.012655449099838734,
-0.006949540227651596,
-0.027402514591813087,
-0.048298370093107224,
-0.04228247329592705,
0.0005372691084630787,
-0.04639260470867157,
... |
154 | AeroAgent: A Vision-Physics-Decision Framework for Aerodynamic Vehicle Design | [
"Ye Liu",
"Shouyi Liu",
"Huiyu Yang",
"Jianghang Gu",
"Wenhao Fan",
"Zhongxin Yang",
"Ding Wang",
"Simeng Chen",
"Zirun Jiang",
"Yuanwei Bin",
"Shiyi Chen",
"Yuntian Chen"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Liu_AeroAgent_A_Vision-Physics-Decision_Framework_for_Aerodynamic_Vehicle_Design_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Liu_AeroAgent_A_Vision-Physics-Decision_Framework_for_Aerodynamic_Vehicle_Design_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Liu_AeroAgent_A_Vision-Physics-Decision_CVPR_2026_supplemental.zip | null | null | @InProceedings{Liu_2026_CVPR,
author = {Liu, Ye and Liu, Shouyi and Yang, Huiyu and Gu, Jianghang and Fan, Wenhao and Yang, Zhongxin and Wang, Ding and Chen, Simeng and Jiang, Zirun and Bin, Yuanwei and Chen, Shiyi and Chen, Yuntian},
title = {AeroAgent: A Vision-Physics-Decision Framework for Aerodynami... | Modern generative models can propose striking 3D vehicle shapes from text and images, but turning these sketches into aerodynamically efficient, regulation compliant designs still requires weeks of high-fidelity computational fluid dynamics (CFD) and manual iteration. As a result, fast 3D generation without trustworthy... | [
0.0010881879134103656,
0.01570359617471695,
0.004521722439676523,
0.0520281083881855,
0.03581694886088371,
0.016488034278154373,
0.0017645173938944936,
-0.005531962029635906,
0.013407133519649506,
-0.05208102986216545,
-0.022423291578888893,
-0.010693872347474098,
-0.048115506768226624,
0.... |
155 | Reviving ConvNeXt for Efficient Convolutional Diffusion Models | [
"Taesung Kwon",
"Lorenzo Bianchi",
"Lennart Wittke",
"Felix Watine",
"Fabio Carrara",
"Jong Chul Ye",
"Romann Weber",
"Vinicius Azevedo"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Kwon_Reviving_ConvNeXt_for_Efficient_Convolutional_Diffusion_Models_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Kwon_Reviving_ConvNeXt_for_Efficient_Convolutional_Diffusion_Models_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Kwon_Reviving_ConvNeXt_for_CVPR_2026_supplemental.pdf | 2603.09408 | cvf | @InProceedings{Kwon_2026_CVPR,
author = {Kwon, Taesung and Bianchi, Lorenzo and Wittke, Lennart and Watine, Felix and Carrara, Fabio and Ye, Jong Chul and Weber, Romann and Azevedo, Vinicius},
title = {Reviving ConvNeXt for Efficient Convolutional Diffusion Models},
booktitle = {Proceedings of the IE... | Recent diffusion models increasingly favor Transformer backbones, motivated by the remarkable scalability of fully attentional architectures. Yet the locality bias, parameter efficiency, and hardware friendliness--the attributes that established ConvNets as the efficient vision backbone--have seen limited exploration i... | [
-0.0018228364642709494,
-0.048324283212423325,
-0.0005403971881605685,
0.06604432314634323,
0.028204388916492462,
0.028506509959697723,
-0.0007814730051904917,
0.029168736189603806,
0.011837407015264034,
-0.038293275982141495,
-0.003150613745674491,
-0.0028277526143938303,
-0.040542855858802... |
156 | StreamAvatar: Streaming Diffusion Models for Real-Time Interactive Human Avatars | [
"Zhiyao Sun",
"Ziqiao Peng",
"Yifeng Ma",
"Yi Chen",
"Zhengguang Zhou",
"Zixiang Zhou",
"Guozhen Zhang",
"Youliang Zhang",
"Yuan Zhou",
"Qinglin Lu",
"Yong-Jin Liu"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Sun_StreamAvatar_Streaming_Diffusion_Models_for_Real-Time_Interactive_Human_Avatars_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Sun_StreamAvatar_Streaming_Diffusion_Models_for_Real-Time_Interactive_Human_Avatars_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Sun_StreamAvatar_Streaming_Diffusion_CVPR_2026_supplemental.zip | 2512.22065 | cvf | @InProceedings{Sun_2026_CVPR,
author = {Sun, Zhiyao and Peng, Ziqiao and Ma, Yifeng and Chen, Yi and Zhou, Zhengguang and Zhou, Zixiang and Zhang, Guozhen and Zhang, Youliang and Zhou, Yuan and Lu, Qinglin and Liu, Yong-Jin},
title = {StreamAvatar: Streaming Diffusion Models for Real-Time Interactive Hum... | Real-time, streaming interactive avatars represent a critical yet challenging goal in digital human research. Although diffusion-based human avatar generation methods achieve remarkable success, their non-causal architecture and high computational costs make them unsuitable for streaming. Moreover, existing interactive... | [
0.034117165952920914,
-0.02502862550318241,
-0.007833662442862988,
0.045984044671058655,
0.0186048224568367,
0.034549862146377563,
0.0399036668241024,
0.019254814833402634,
-0.017372997477650642,
-0.05867693945765495,
-0.05084484815597534,
-0.023321600630879402,
-0.06757233291864395,
-0.01... |
157 | DetAny4D: Detect Anything 4D Temporally in a Streaming RGB Video | [
"Jiawei Hou",
"Shenghao Zhang",
"Can Wang",
"Zheng Gu",
"Yonggen Ling",
"Taiping Zeng",
"Xiangyang Xue",
"Jingbo Zhang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Hou_DetAny4D_Detect_Anything_4D_Temporally_in_a_Streaming_RGB_Video_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Hou_DetAny4D_Detect_Anything_4D_Temporally_in_a_Streaming_RGB_Video_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Hou_DetAny4D_Detect_Anything_CVPR_2026_supplemental.zip | 2511.18814 | cvf | @InProceedings{Hou_2026_CVPR,
author = {Hou, Jiawei and Zhang, Shenghao and Wang, Can and Gu, Zheng and Ling, Yonggen and Zeng, Taiping and Xue, Xiangyang and Zhang, Jingbo},
title = {DetAny4D: Detect Anything 4D Temporally in a Streaming RGB Video},
booktitle = {Proceedings of the IEEE/CVF Conferenc... | Reliable 4D object detection, which refers to 3D object detection in streaming video, is crucial for perceiving and understanding the real world. Existing open-set 4D object detection methods typically make predictions on a frame-by-frame basis without modeling temporal consistency, or rely on complex multi-stage pipel... | [
0.006678307428956032,
0.0044533018954098225,
0.012322904542088509,
0.05406299605965614,
0.0007665462326258421,
0.015002458356320858,
0.01946640945971012,
0.026264775544404984,
-0.030812814831733704,
-0.03352822735905647,
-0.020605547353625298,
-0.012562872841954231,
-0.06903236359357834,
0... |
158 | Rethinking Token Reduction for Large Vision-Language Models | [
"Yi Wang",
"Haofei Zhang",
"Qihan Huang",
"Anda Cao",
"Gongfan Fang",
"Wei Wang",
"Xuan Jin",
"Jie Song",
"Mingli Song",
"Xinchao Wang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Wang_Rethinking_Token_Reduction_for_Large_Vision-Language_Models_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Wang_Rethinking_Token_Reduction_for_Large_Vision-Language_Models_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Wang_Rethinking_Token_Reduction_CVPR_2026_supplemental.pdf | 2603.21701 | cvf | @InProceedings{Wang_2026_CVPR,
author = {Wang, Yi and Zhang, Haofei and Huang, Qihan and Cao, Anda and Fang, Gongfan and Wang, Wei and Jin, Xuan and Song, Jie and Song, Mingli and Wang, Xinchao},
title = {Rethinking Token Reduction for Large Vision-Language Models},
booktitle = {Proceedings of the IE... | Large Vision-Language Models (LVLMs) excel in visual understanding and reasoning, but the excessive visual tokens lead to high inference costs. Although recent token reduction methods mitigate this issue, they mainly target single-turn Visual Question Answering (VQA), leaving the more practical multi-turn VQA (MT-VQA) ... | [
0.01711416430771351,
-0.014148455113172531,
0.010381216183304787,
0.04394961893558502,
0.0075134471990168095,
0.05054730921983719,
0.012235956266522408,
0.016935858875513077,
-0.021088819950819016,
-0.012200113385915756,
-0.04487377032637596,
0.03358536213636398,
-0.06072121113538742,
-0.0... |
159 | CaptionQA: Is Your Caption as Useful as the Image Itself? | [
"Shijia Yang",
"Yunong Liu",
"Bohan Zhai",
"Ximeng Sun",
"Zicheng Liu",
"Emad Barsoum",
"Manling Li",
"Chenfeng Xu"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Yang_CaptionQA_Is_Your_Caption_as_Useful_as_the_Image_Itself_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Yang_CaptionQA_Is_Your_Caption_as_Useful_as_the_Image_Itself_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Yang_CaptionQA_Is_Your_CVPR_2026_supplemental.pdf | 2511.21025 | cvf | @InProceedings{Yang_2026_CVPR,
author = {Yang, Shijia and Liu, Yunong and Zhai, Bohan and Sun, Ximeng and Liu, Zicheng and Barsoum, Emad and Li, Manling and Xu, Chenfeng},
title = {CaptionQA: Is Your Caption as Useful as the Image Itself?},
booktitle = {Proceedings of the IEEE/CVF Conference on Compu... | Image captions serve as efficient surrogates for visual content in multimodal systems such as retrieval, recommendation, multi-step agentic inference pipelines. Yet current evaluation practices miss a fundamental question: Can captions stand-in for images in real downstream tasks? We propose a utility-based benchmark, ... | [
-0.001601237803697586,
-0.04269269481301308,
0.0036057785619050264,
0.0754283219575882,
0.033568739891052246,
-0.016654780134558678,
0.012401125393807888,
0.004036791156977415,
-0.01905667595565319,
-0.0015143903438001871,
-0.05474357679486275,
0.06751158833503723,
-0.06719427555799484,
-0... |
160 | Dynamics-Aware Preference Optimization for Vision-Language Models | [
"Jusheng Zhang",
"Kaitong Cai",
"Jing Yang",
"Jian Wang",
"Keze Wang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Zhang_Dynamics-Aware_Preference_Optimization_for_Vision-Language_Models_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Zhang_Dynamics-Aware_Preference_Optimization_for_Vision-Language_Models_CVPR_2026_paper.pdf | null | null | null | @InProceedings{Zhang_2026_CVPR,
author = {Zhang, Jusheng and Cai, Kaitong and Yang, Jing and Wang, Jian and Wang, Keze},
title = {Dynamics-Aware Preference Optimization for Vision-Language Models},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},... | Preference-based finetuning of vision-language models (VLMs) is notoriously unstable, as trivially wrong negatives inject uninformative gradients that distort optimization and degrade calibration. This work revisits this issue through the lens of learning dynamics and identifies a core pathology, the squeezing effect, ... | [
-0.013326812535524368,
0.008756984025239944,
0.04765008017420769,
0.03088039718568325,
0.023197177797555923,
0.054586976766586304,
0.014659928157925606,
0.016909785568714142,
-0.03436706215143204,
-0.03762136772274971,
-0.026609128341078758,
0.04419249668717384,
-0.07020179182291031,
-0.00... |
161 | Hierarchical Long Video Understanding with Audiovisual Entity Cohesion and Agentic Search | [
"Xinlei Yin",
"Xiulian Peng",
"Xiao Li",
"Zhiwei Xiong",
"Yan Lu"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Yin_Hierarchical_Long_Video_Understanding_with_Audiovisual_Entity_Cohesion_and_Agentic_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Yin_Hierarchical_Long_Video_Understanding_with_Audiovisual_Entity_Cohesion_and_Agentic_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Yin_Hierarchical_Long_Video_CVPR_2026_supplemental.pdf | 2601.13719 | cvf | @InProceedings{Yin_2026_CVPR,
author = {Yin, Xinlei and Peng, Xiulian and Li, Xiao and Xiong, Zhiwei and Lu, Yan},
title = {Hierarchical Long Video Understanding with Audiovisual Entity Cohesion and Agentic Search},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Re... | Long video understanding presents significant challenges for vision-language models due to extremely long context windows. Existing solutions relying on naive chunking strategies with retrieval-augmented generation, typically suffer from information fragmentation and a loss of global coherence. We present HAVEN, a unif... | [
-0.003864818252623081,
0.014720411039888859,
0.015491105616092682,
0.03085831180214882,
0.03225899487733841,
0.00152226269710809,
0.03122374601662159,
0.011228997260332108,
-0.0409620925784111,
-0.011541259475052357,
-0.032360028475522995,
0.018329614773392677,
-0.06264002621173859,
-0.000... |
162 | FlowDC: Flow-Based Decoupling-Decay for Complex Image Editing | [
"Yilei Jiang",
"Zhen Wang",
"Yanghao Wang",
"Jun Yu",
"Yueting Zhuang",
"Jun Xiao",
"Long Chen"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Jiang_FlowDC_Flow-Based_Decoupling-Decay_for_Complex_Image_Editing_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Jiang_FlowDC_Flow-Based_Decoupling-Decay_for_Complex_Image_Editing_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Jiang_FlowDC_Flow-Based_Decoupling-Decay_CVPR_2026_supplemental.pdf | 2512.11395 | cvf | @InProceedings{Jiang_2026_CVPR,
author = {Jiang, Yilei and Wang, Zhen and Wang, Yanghao and Yu, Jun and Zhuang, Yueting and Xiao, Jun and Chen, Long},
title = {FlowDC: Flow-Based Decoupling-Decay for Complex Image Editing},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pa... | With the surge of pre-trained text-to-image flow matching models, text-based image editing performance has gained remarkable improvement, especially for **simple editing** that only contains a single editing target. However, to satisfy the exploding editing requirements, the **complex editing** that contains multiple e... | [
0.0006706343847326934,
-0.009615778923034668,
0.005528839770704508,
0.06597266346216202,
0.0438387393951416,
0.04284132272005081,
0.01800672337412834,
0.007165917661041021,
-0.01680212840437889,
-0.0559956356883049,
-0.008481929078698158,
-0.0049088080413639545,
-0.058790743350982666,
-0.0... |
163 | PDCR: Perception-Decomposed Confidence Reward for Vision-Language Reasoning | [
"Hee Suk Yoon",
"Eunseop Yoon",
"Ji Woo Hong",
"SooHwan Eom",
"Gwanhyeong Koo",
"Mark Hasegawa-Johnson",
"Qi Dai",
"Chong Luo",
"Chang D. Yoo"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Yoon_PDCR_Perception-Decomposed_Confidence_Reward_for_Vision-Language_Reasoning_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Yoon_PDCR_Perception-Decomposed_Confidence_Reward_for_Vision-Language_Reasoning_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Yoon_PDCR_Perception-Decomposed_Confidence_CVPR_2026_supplemental.pdf | 2605.13467 | cvf | @InProceedings{Yoon_2026_CVPR,
author = {Yoon, Hee Suk and Yoon, Eunseop and Hong, Ji Woo and Eom, SooHwan and Koo, Gwanhyeong and Hasegawa-Johnson, Mark and Dai, Qi and Luo, Chong and Yoo, Chang D.},
title = {PDCR: Perception-Decomposed Confidence Reward for Vision-Language Reasoning},
booktitle = {... | Reinforcement Learning with Verifiable Rewards (RLVR) traditionally relies on a sparse, outcome-based signal. Recent work shows that providing a fine-grained, model-intrinsic signal--rewarding the confidence growth in the ground-truth answer--effectively improves language reasoning training by providing step-level guid... | [
-0.021490389481186867,
0.013150248676538467,
-0.0021972921676933765,
0.07918494194746017,
0.01171619351953268,
0.048867467790842056,
0.01865711621940136,
0.011166103184223175,
-0.0323653370141983,
-0.031197864562273026,
-0.04613300785422325,
0.04814256727695465,
-0.05519714578986168,
-0.00... |
164 | CaptionFormer: Unified Segmentation, Tracking, and Captioning for Spatio-Temporal Objects | [
"Gabriel Fiastre",
"Antoine Yang",
"Cordelia Schmid"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Fiastre_CaptionFormer_Unified_Segmentation_Tracking_and_Captioning_for_Spatio-Temporal_Objects_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Fiastre_CaptionFormer_Unified_Segmentation_Tracking_and_Captioning_for_Spatio-Temporal_Objects_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Fiastre_CaptionFormer_Unified_Segmentation_CVPR_2026_supplemental.pdf | null | null | @InProceedings{Fiastre_2026_CVPR,
author = {Fiastre, Gabriel and Yang, Antoine and Schmid, Cordelia},
title = {CaptionFormer: Unified Segmentation, Tracking, and Captioning for Spatio-Temporal Objects},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CV... | Dense Video Object Captioning (DVOC) is the task of jointly detecting, tracking, and captioning object trajectories in a video, requiring the ability to understand spatio-temporal details and describe them in natural language.Due to the complexity of the task and the high cost associated with manual annotation, previou... | [
0.007834222167730331,
-0.03961788862943649,
0.03386899083852768,
0.03462565317749977,
0.005533058196306229,
0.017714807763695717,
0.0076651424169540405,
0.03136594593524933,
-0.04010166972875595,
-0.042214032262563705,
-0.04765631631016731,
-0.006947437301278114,
-0.05182706564664841,
0.01... |
165 | MeanFlow Transformers with Representation Autoencoders | [
"Zheyuan Hu",
"Chieh-Hsin Lai",
"Ge Wu",
"Yuki Mitsufuji",
"Stefano Ermon"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Hu_MeanFlow_Transformers_with_Representation_Autoencoders_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Hu_MeanFlow_Transformers_with_Representation_Autoencoders_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Hu_MeanFlow_Transformers_with_CVPR_2026_supplemental.pdf | 2511.13019 | cvf | @InProceedings{Hu_2026_CVPR,
author = {Hu, Zheyuan and Lai, Chieh-Hsin and Wu, Ge and Mitsufuji, Yuki and Ermon, Stefano},
title = {MeanFlow Transformers with Representation Autoencoders},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
mont... | MeanFlow (MF) is a diffusion-motivated generative model that enables efficient few-step generation by learning long jumps directly from noise to data. In practice, it is often used as a latent MF by leveraging the pre-trained Stable Diffusion variational autoencoder (SD-VAE) for high-dimensional data modeling. However,... | [
0.00021106601343490183,
-0.022270549088716507,
-0.009564981795847416,
0.0561397485435009,
0.034599412232637405,
0.05898800492286682,
0.03559866175055504,
-0.014018844813108444,
-0.019963258877396584,
-0.05688623711466789,
0.0004943655803799629,
-0.027279848232865334,
-0.0474984310567379,
0... |
166 | WeMMU: Enhanced Bridging of Vision-Language Models and Diffusion Models via Noisy Query Tokens | [
"Jian Yang",
"Dacheng Yin",
"Xiaoxuan He",
"Yong Li",
"Fengyun Rao",
"Jing Lyu",
"Wei Zhai",
"Yang Cao",
"Zheng-Jun Zha"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Yang_WeMMU_Enhanced_Bridging_of_Vision-Language_Models_and_Diffusion_Models_via_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Yang_WeMMU_Enhanced_Bridging_of_Vision-Language_Models_and_Diffusion_Models_via_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Yang_WeMMU_Enhanced_Bridging_CVPR_2026_supplemental.pdf | 2512.02536 | cvf | @InProceedings{Yang_2026_CVPR,
author = {Yang, Jian and Yin, Dacheng and He, Xiaoxuan and Li, Yong and Rao, Fengyun and Lyu, Jing and Zhai, Wei and Cao, Yang and Zha, Zheng-Jun},
title = {WeMMU: Enhanced Bridging of Vision-Language Models and Diffusion Models via Noisy Query Tokens},
booktitle = {Pro... | Recent progress in multimodal large language models (MLLMs) has highlighted the challenge of efficiently bridging pre-trained Vision-Language Models (VLMs) with Diffusion Models. While methods using a fixed number of learnable query tokens offer computational efficiency, they suffer from task generalization collapse, f... | [
0.028962040320038795,
-0.0047113834880292416,
0.005829050671309233,
0.04025408253073692,
-0.007556100841611624,
0.04866686835885048,
0.048523496836423874,
0.01123450044542551,
-0.027471810579299927,
-0.044010717421770096,
-0.03328496217727661,
0.017229659482836723,
-0.058018848299980164,
0... |
167 | PAM: A Pose-Appearance-Motion Engine for Sim-to-Real HOI Video Generation | [
"Mingju Gao",
"Kaisen Yang",
"Huan-ang Gao",
"Bohan Li",
"Ao Ding",
"Wenyi Li",
"Yangcheng Yu",
"Jinkun Liu",
"Shaocong Xu",
"Yike Niu",
"Haohan Chi",
"Hao Chen",
"Hao Tang",
"Yu Zhang",
"Li Yi",
"Hao Zhao"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Gao_PAM_A_Pose-Appearance-Motion_Engine_for_Sim-to-Real_HOI_Video_Generation_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Gao_PAM_A_Pose-Appearance-Motion_Engine_for_Sim-to-Real_HOI_Video_Generation_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Gao_PAM_A_Pose-Appearance-Motion_CVPR_2026_supplemental.pdf | 2603.22193 | cvf | @InProceedings{Gao_2026_CVPR,
author = {Gao, Mingju and Yang, Kaisen and Gao, Huan-ang and Li, Bohan and Ding, Ao and Li, Wenyi and Yu, Yangcheng and Liu, Jinkun and Xu, Shaocong and Niu, Yike and Chi, Haohan and Chen, Hao and Tang, Hao and Zhang, Yu and Yi, Li and Zhao, Hao},
title = {PAM: A Pose-Appear... | Hand-object interaction (HOI) reconstruction and synthesis are becoming central to embodied AI and AR/VR. Yet, despite rapid progress, existing HOI generation research remains fragmented across three disjoint tracks: (1) pose-only synthesis that predicts MANO trajectories without producing pixels; (2) single-image HOI ... | [
-0.022196635603904724,
0.02542249672114849,
-0.010218553245067596,
0.026468267664313316,
0.02474362961947918,
0.04247551038861275,
0.03252031281590462,
0.013914313167333603,
-0.04196719825267792,
-0.0392594113945961,
-0.017082661390304565,
-0.013955029658973217,
-0.07965592294931412,
-0.00... |
168 | CLAY: Conditional Visual Similarity Modulation in Vision-Language Embedding Space | [
"Sohwi Lim",
"Lee Hyoseok",
"Jungjoon Park",
"Tae-Hyun Oh"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Lim_CLAY_Conditional_Visual_Similarity_Modulation_in_Vision-Language_Embedding_Space_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Lim_CLAY_Conditional_Visual_Similarity_Modulation_in_Vision-Language_Embedding_Space_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Lim_CLAY_Conditional_Visual_CVPR_2026_supplemental.pdf | 2604.11539 | cvf | @InProceedings{Lim_2026_CVPR,
author = {Lim, Sohwi and Hyoseok, Lee and Park, Jungjoon and Oh, Tae-Hyun},
title = {CLAY: Conditional Visual Similarity Modulation in Vision-Language Embedding Space},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}... | Human perception of visual similarity is inherently adaptive and subjective, depending on the users' interests and focus. However, most image retrieval systems fail to reflect this flexibility, relying on a fixed, monolithic metric that cannot incorporate multiple conditions simultaneously. To address this, we propose ... | [
0.01059955544769764,
-0.014944477006793022,
-0.010115447454154491,
0.021995829418301582,
0.048226047307252884,
0.02667240984737873,
0.004408267792314291,
0.015298455953598022,
-0.029594821855425835,
-0.035111572593450546,
-0.058297038078308105,
0.029866104945540428,
-0.04906725510954857,
-... |
169 | The Surprising Effectiveness of Noise Pretraining for Implicit Neural Representations | [
"Kushal Vyas",
"Alper Kayabasi",
"Daniel Kim",
"Vishwanath Saragadam",
"Ashok Veeraraghavan",
"Guha Balakrishnan"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Vyas_The_Surprising_Effectiveness_of_Noise_Pretraining_for_Implicit_Neural_Representations_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Vyas_The_Surprising_Effectiveness_of_Noise_Pretraining_for_Implicit_Neural_Representations_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Vyas_The_Surprising_Effectiveness_CVPR_2026_supplemental.pdf | 2603.29034 | cvf | @InProceedings{Vyas_2026_CVPR,
author = {Vyas, Kushal and Kayabasi, Alper and Kim, Daniel and Saragadam, Vishwanath and Veeraraghavan, Ashok and Balakrishnan, Guha},
title = {The Surprising Effectiveness of Noise Pretraining for Implicit Neural Representations},
booktitle = {Proceedings of the IEEE/C... | The approximation and convergence properties of implicit neural representations (INRs) are known to be highly sensitive to parameter initialization strategies. While several data-driven initialization methods demonstrate significant improvements over standard random sampling, the reasons for their success -- specifical... | [
-0.0076978388242423534,
-0.004625031724572182,
-0.006893981248140335,
0.04106086492538452,
0.03023829124867916,
0.03230617940425873,
0.03918365389108658,
0.0006601472850888968,
-0.0335376113653183,
-0.05572590231895447,
-0.016085371375083923,
-0.007418021559715271,
-0.05352434515953064,
-0... |
170 | Learning from Semantic Dictionaries: Discriminative Codebook Contrastive Learning for Unified Visual Representation and Generation | [
"Imanol G. Estepa",
"Jesús M. Rodríguez-de-Vera",
"Bhalaji Nagarajan",
"Petia Radeva"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Estepa_Learning_from_Semantic_Dictionaries_Discriminative_Codebook_Contrastive_Learning_for_Unified_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Estepa_Learning_from_Semantic_Dictionaries_Discriminative_Codebook_Contrastive_Learning_for_Unified_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Estepa_Learning_from_Semantic_CVPR_2026_supplemental.pdf | 2605.25012 | title_snapshot | @InProceedings{Estepa_2026_CVPR,
author = {Estepa, Imanol G. and Rodr{\'\i}guez-de-Vera, Jes\'us M. and Nagarajan, Bhalaji and Radeva, Petia},
title = {Learning from Semantic Dictionaries: Discriminative Codebook Contrastive Learning for Unified Visual Representation and Generation},
booktitle = {Pro... | Discriminative and generative vision models excel in their respective domains but remain semantically misaligned, hindering progress toward unified visual learning. We introduce LEASE (LEArning from SEmantic Dictionaries), a self-supervised framework that bridges this gap using a paired generative-discriminative codebo... | [
0.023211656138300896,
-0.03385363519191742,
-0.008385817520320415,
0.06662800908088684,
0.029659118503332138,
0.03247375041246414,
0.007764862850308418,
0.004863697104156017,
-0.013874905183911324,
-0.02700573205947876,
-0.05651122331619263,
0.0008003454422578216,
-0.0880056694149971,
0.01... |
171 | ZipMap: Linear-Time Stateful 3D Reconstruction via Test-Time Training | [
"Haian Jin",
"Rundi Wu",
"Tianyuan Zhang",
"Ruiqi Gao",
"Jonathan T. Barron",
"Noah Snavely",
"Aleksander Hołyński"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Jin_ZipMap_Linear-Time_Stateful_3D_Reconstruction_via_Test-Time_Training_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Jin_ZipMap_Linear-Time_Stateful_3D_Reconstruction_via_Test-Time_Training_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Jin_ZipMap_Linear-Time_Stateful_CVPR_2026_supplemental.pdf | 2603.04385 | cvf | @InProceedings{Jin_2026_CVPR,
author = {Jin, Haian and Wu, Rundi and Zhang, Tianyuan and Gao, Ruiqi and Barron, Jonathan T. and Snavely, Noah and Ho{\l}y\'nski, Aleksander},
title = {ZipMap: Linear-Time Stateful 3D Reconstruction via Test-Time Training},
booktitle = {Proceedings of the IEEE/CVF Confe... | Feed-forward transformer models have driven rapid progress in 3D vision, but state-of-the-art methods such as VGGT and \pi^3 have a computational cost that scales quadratically with the number of input images, making them inefficient when applied to large image collections. Sequential-reconstruction approaches reduce t... | [
0.030178586021065712,
-0.029268478974699974,
0.0013457642635330558,
0.02308954857289791,
0.03551597148180008,
0.052987176924943924,
-0.008388801477849483,
0.025727251544594765,
-0.008473044261336327,
-0.0667009949684143,
-0.03914971649646759,
-0.02076520211994648,
-0.07269114255905151,
0.0... |
172 | SGDrive: Scene-to-Goal Hierarchical World Cognition for Autonomous Driving | [
"Jingyu Li",
"Junjie Wu",
"Dongnan Hu",
"Xiangkai Huang",
"Bin Sun",
"Zhihui Hao",
"Xianpeng Lang",
"Xiatian Zhu",
"Li Zhang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Li_SGDrive_Scene-to-Goal_Hierarchical_World_Cognition_for_Autonomous_Driving_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Li_SGDrive_Scene-to-Goal_Hierarchical_World_Cognition_for_Autonomous_Driving_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Li_SGDrive_Scene-to-Goal_Hierarchical_CVPR_2026_supplemental.pdf | 2601.05640 | cvf | @InProceedings{Li_2026_CVPR,
author = {Li, Jingyu and Wu, Junjie and Hu, Dongnan and Huang, Xiangkai and Sun, Bin and Hao, Zhihui and Lang, Xianpeng and Zhu, Xiatian and Zhang, Li},
title = {SGDrive: Scene-to-Goal Hierarchical World Cognition for Autonomous Driving},
booktitle = {Proceedings of the I... | Recent end-to-end autonomous driving approaches have leveraged Vision-Language Models (VLMs) to enhance planning capabilities in complex driving scenarios. However, VLMs are inherently trained as generalist models, lacking specialized understanding of driving-specific reasoning in 3D space and time. When applied to aut... | [
-0.013791547156870365,
0.016222521662712097,
0.004917509388178587,
0.038709912449121475,
0.03330611810088158,
0.02284020185470581,
0.04327704384922981,
0.031105924397706985,
-0.0081532783806324,
-0.03691331669688225,
-0.01857156492769718,
0.01447695679962635,
-0.06709454953670502,
-0.02650... |
173 | OnlinePG: Online Open-Vocabulary Panoptic Mapping with 3D Gaussian Splatting | [
"Hongjia Zhai",
"Qi Zhang",
"Xiaokun Pan",
"Xiyu Zhang",
"Yitong Dong",
"Huaqi Zhang",
"Dan Xu",
"Guofeng Zhang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Zhai_OnlinePG_Online_Open-Vocabulary_Panoptic_Mapping_with_3D_Gaussian_Splatting_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Zhai_OnlinePG_Online_Open-Vocabulary_Panoptic_Mapping_with_3D_Gaussian_Splatting_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zhai_OnlinePG_Online_Open-Vocabulary_CVPR_2026_supplemental.pdf | 2603.18510 | cvf | @InProceedings{Zhai_2026_CVPR,
author = {Zhai, Hongjia and Zhang, Qi and Pan, Xiaokun and Zhang, Xiyu and Dong, Yitong and Zhang, Huaqi and Xu, Dan and Zhang, Guofeng},
title = {OnlinePG: Online Open-Vocabulary Panoptic Mapping with 3D Gaussian Splatting},
booktitle = {Proceedings of the IEEE/CVF Con... | Open-vocabulary scene understanding with online panoptic mapping is essential for embodied applications to perceive and interact with environments. However, existing methods are predominantly offline or lack instance-level understanding, limiting their applicability to real-world robotic tasks. In this paper, we propos... | [
-0.0028410302475094795,
0.009577828459441662,
0.02856859564781189,
0.03521748632192612,
0.01649423874914646,
0.058003466576337814,
0.017893049865961075,
0.024416182190179825,
-0.021463431417942047,
-0.03652017191052437,
-0.03226625174283981,
-0.011041372083127499,
-0.06401987373828888,
-0.... |
174 | Are Image-to-Video Models Good Zero-Shot Image Editors? | [
"Zechuan Zhang",
"Zhenyuan Chen",
"Zongxin Yang",
"Yi Yang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Zhang_Are_Image-to-Video_Models_Good_Zero-Shot_Image_Editors_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Zhang_Are_Image-to-Video_Models_Good_Zero-Shot_Image_Editors_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zhang_Are_Image-to-Video_Models_CVPR_2026_supplemental.pdf | 2511.19435 | cvf | @InProceedings{Zhang_2026_CVPR,
author = {Zhang, Zechuan and Chen, Zhenyuan and Yang, Zongxin and Yang, Yi},
title = {Are Image-to-Video Models Good Zero-Shot Image Editors?},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June... | Large-scale video diffusion models exhibit strong world-simulation and temporal reasoning capabilities, yet their potential as zero-shot image editors remains underexplored. We present \ifedit IF-Edit (Image Edit by Generating Frames), a tuning-free framework that repurposes pre-trained image-to-video diffusion models... | [
0.028539296239614487,
-0.005102967377752066,
-0.002162050688639283,
0.07234525680541992,
0.04995310679078102,
0.02216191776096821,
0.015229552052915096,
0.03331175446510315,
-0.023994971066713333,
-0.043827079236507416,
-0.033604443073272705,
-0.006582047324627638,
-0.0323893241584301,
-0.... |
175 | PyraTok: Language-Aligned Pyramidal Tokenizer for Video Understanding and Generation | [
"Onkar Susladkar",
"Tushar Prakash",
"Adheesh Juvekar",
"Kiet A. Nguyen",
"Dong-Hwan Jang",
"Inderjit S Dhillon",
"Ismini Lourentzou"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Susladkar_PyraTok_Language-Aligned_Pyramidal_Tokenizer_for_Video_Understanding_and_Generation_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Susladkar_PyraTok_Language-Aligned_Pyramidal_Tokenizer_for_Video_Understanding_and_Generation_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Susladkar_PyraTok_Language-Aligned_Pyramidal_CVPR_2026_supplemental.pdf | 2601.16210 | cvf | @InProceedings{Susladkar_2026_CVPR,
author = {Susladkar, Onkar and Prakash, Tushar and Juvekar, Adheesh and Nguyen, Kiet A. and Jang, Dong-Hwan and Dhillon, Inderjit S and Lourentzou, Ismini},
title = {PyraTok: Language-Aligned Pyramidal Tokenizer for Video Understanding and Generation},
booktitle = ... | Discrete video VAEs underpin modern text-to-video generation and video understanding systems, yet existing tokenizers typically learn visual codebooks at a single scale with limited vocabularies and shallow language supervision, leading to poor cross-modal alignment and zero-shot transfer. We introduce PyraTok, a langu... | [
0.03131765127182007,
-0.017626089975237846,
0.00696510449051857,
0.060736242681741714,
-0.013630754314363003,
0.08008257299661636,
0.02888057939708233,
0.015345657244324684,
-0.04004955664277077,
-0.018027029931545258,
-0.018921278417110443,
-0.014471660368144512,
-0.044253911823034286,
0.... |
176 | RT-Splatting: Joint Reflection-Transmission Modeling with Gaussian Splatting | [
"Ji Shi",
"Xianghua Ying",
"Bowei Xing",
"Ruohao Guo",
"Wenzhen Yue"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Shi_RT-Splatting_Joint_Reflection-Transmission_Modeling_with_Gaussian_Splatting_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Shi_RT-Splatting_Joint_Reflection-Transmission_Modeling_with_Gaussian_Splatting_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Shi_RT-Splatting_Joint_Reflection-Transmission_CVPR_2026_supplemental.pdf | 2605.18263 | cvf | @InProceedings{Shi_2026_CVPR,
author = {Shi, Ji and Ying, Xianghua and Xing, Bowei and Guo, Ruohao and Yue, Wenzhen},
title = {RT-Splatting: Joint Reflection-Transmission Modeling with Gaussian Splatting},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition ... | 3D Gaussian Splatting (3DGS) enables real-time novel view synthesis with high visual quality. However, existing methods struggle with semi-transparent specular surfaces that exhibit both complex reflections and clear transmission, often producing blurry reflections or overly occluded transmission. To address this, we p... | [
0.034457236528396606,
0.00492269080132246,
0.04952283948659897,
0.03193870186805725,
0.004331408534198999,
0.009781928732991219,
0.009954792447388172,
0.027564529329538345,
-0.011495758779346943,
-0.08121941983699799,
-0.01813367009162903,
-0.005630426574498415,
-0.055552974343299866,
0.01... |
177 | From Observation to Action: Latent Action-based Primitive Segmentation for VLA Pre-training in Industrial Settings | [
"Jiajie Zhang",
"Sören Schwertfeger",
"Alexander Kleiner"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Zhang_From_Observation_to_Action_Latent_Action-based_Primitive_Segmentation_for_VLA_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Zhang_From_Observation_to_Action_Latent_Action-based_Primitive_Segmentation_for_VLA_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zhang_From_Observation_to_CVPR_2026_supplemental.pdf | 2511.21428 | cvf | @InProceedings{Zhang_2026_CVPR,
author = {Zhang, Jiajie and Schwertfeger, S\"oren and Kleiner, Alexander},
title = {From Observation to Action: Latent Action-based Primitive Segmentation for VLA Pre-training in Industrial Settings},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Visi... | We present a novel unsupervised framework to unlock vast unlabeled human demonstration data from continuous industrial video streams for Vision-Language-Action (VLA) model pre-training. Our method first trains a lightweight motion tokenizer to encode motion dynamics, then employs an unsupervised action segmenter levera... | [
0.02770007960498333,
-0.023637892678380013,
-0.028428543359041214,
0.016661511734128,
0.028117740526795387,
0.02274179272353649,
0.04702743887901306,
0.027137670665979385,
-0.03595263510942459,
-0.03216918930411339,
-0.04078322649002075,
-0.012539767660200596,
-0.050543345510959625,
-0.011... |
178 | Retrieve and Segment: Are a Few Examples Enough to Bridge the Supervision Gap in Open-Vocabulary Segmentation? | [
"Tilemachos Aravanis",
"Vladan Stojnić",
"Bill Psomas",
"Nikos Komodakis",
"Giorgos Tolias"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Aravanis_Retrieve_and_Segment_Are_a_Few_Examples_Enough_to_Bridge_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Aravanis_Retrieve_and_Segment_Are_a_Few_Examples_Enough_to_Bridge_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Aravanis_Retrieve_and_Segment_CVPR_2026_supplemental.pdf | 2602.23339 | title_snapshot | @InProceedings{Aravanis_2026_CVPR,
author = {Aravanis, Tilemachos and Stojni\'c, Vladan and Psomas, Bill and Komodakis, Nikos and Tolias, Giorgos},
title = {Retrieve and Segment: Are a Few Examples Enough to Bridge the Supervision Gap in Open-Vocabulary Segmentation?},
booktitle = {Proceedings of the... | Open-vocabulary segmentation (OVS) extends the zero-shot recognition capabilities of vision-language models (VLMs) to pixel-level prediction, enabling segmentation of arbitrary categories specified by text prompts. Despite recent progress, OVS lags behind fully supervised approaches due to two challenges: the coarse im... | [
-0.018132947385311127,
-0.026495881378650665,
-0.0024709543213248253,
0.07040907442569733,
0.02296636812388897,
0.009747453033924103,
0.047377388924360275,
0.041772253811359406,
-0.028388790786266327,
0.0070579475723207,
-0.03596419841051102,
0.03865884616971016,
-0.06548041850328445,
-0.0... |
179 | Shoe Style-Invariant and Ground-Aware Learning for Dense Foot Contact Estimation | [
"Daniel Sungho Jung",
"Kyoung Mu Lee"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Jung_Shoe_Style-Invariant_and_Ground-Aware_Learning_for_Dense_Foot_Contact_Estimation_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Jung_Shoe_Style-Invariant_and_Ground-Aware_Learning_for_Dense_Foot_Contact_Estimation_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Jung_Shoe_Style-Invariant_and_CVPR_2026_supplemental.pdf | 2511.22184 | cvf | @InProceedings{Jung_2026_CVPR,
author = {Jung, Daniel Sungho and Lee, Kyoung Mu},
title = {Shoe Style-Invariant and Ground-Aware Learning for Dense Foot Contact Estimation},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},... | Foot contact plays a critical role in human interaction with the world, and thus exploring foot contact can advance our understanding of human movement and physical interaction. Despite its importance, existing methods often approximate foot contact using a zero-velocity constraint and focus on joint-level contact, fai... | [
-0.006797047331929207,
-0.011762050911784172,
-0.0045659225434064865,
0.03257271274924278,
0.03307564929127693,
0.04700002446770668,
0.03844431787729263,
0.003634297288954258,
0.0001950299774762243,
-0.04778357222676277,
-0.04321354627609253,
-0.013184726238250732,
-0.11361327022314072,
-0... |
180 | UniGeoSeg: Towards Unified Open-World Segmentation for Geospatial Scenes | [
"Shuo Ni",
"Di Wang",
"He Chen",
"Haonan Guo",
"Ning Zhang",
"Jing Zhang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Ni_UniGeoSeg_Towards_Unified_Open-World_Segmentation_for_Geospatial_Scenes_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Ni_UniGeoSeg_Towards_Unified_Open-World_Segmentation_for_Geospatial_Scenes_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Ni_UniGeoSeg_Towards_Unified_CVPR_2026_supplemental.pdf | 2511.23332 | cvf | @InProceedings{Ni_2026_CVPR,
author = {Ni, Shuo and Wang, Di and Chen, He and Guo, Haonan and Zhang, Ning and Zhang, Jing},
title = {UniGeoSeg: Towards Unified Open-World Segmentation for Geospatial Scenes},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognitio... | Instruction-driven segmentation in remote sensing generates masks from guidance, offering great potential for accessible and generalizable applications. However, existing methods suffer from fragmented task formulations and limited instruction data, hindering effective understanding and generalization. To address these... | [
0.017131434753537178,
-0.014231483452022076,
-0.005221550352871418,
0.029961343854665756,
0.029590150341391563,
0.01694631762802601,
0.03780115768313408,
0.03187968581914902,
-0.012133966200053692,
-0.03423386439681053,
-0.029508860781788826,
0.00587419280782342,
-0.07597307115793228,
-0.0... |
181 | ShiftLUT: Spatial Shift Enhanced Look-Up Tables for Efficient Image Restoration | [
"Xiaolong Zeng",
"Yitong Yu",
"Shiyao Xiong",
"Jinhua Hao",
"Ming Sun",
"Chao Zhou",
"Bin Wang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Zeng_ShiftLUT_Spatial_Shift_Enhanced_Look-Up_Tables_for_Efficient_Image_Restoration_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Zeng_ShiftLUT_Spatial_Shift_Enhanced_Look-Up_Tables_for_Efficient_Image_Restoration_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zeng_ShiftLUT_Spatial_Shift_CVPR_2026_supplemental.pdf | 2603.00906 | cvf | @InProceedings{Zeng_2026_CVPR,
author = {Zeng, Xiaolong and Yu, Yitong and Xiong, Shiyao and Hao, Jinhua and Sun, Ming and Zhou, Chao and Wang, Bin},
title = {ShiftLUT: Spatial Shift Enhanced Look-Up Tables for Efficient Image Restoration},
booktitle = {Proceedings of the IEEE/CVF Conference on Compu... | Look-Up Table based methods have emerged as a promising direction for efficient image restoration tasks. Recent LUT-based methods focus on improving their performance by expanding the receptive field. However, they inevitably introduce extra computational and storage overhead, which hinders their deployment in edge dev... | [
-0.009982465766370296,
-0.03422541171312332,
0.0004095395270269364,
0.04908440262079239,
0.06866738200187683,
0.04567286744713783,
0.006276802159845829,
0.009538288228213787,
-0.028039434924721718,
-0.060581110417842865,
0.0056970976293087006,
-0.004835138563066721,
-0.039446163922548294,
... |
182 | NaTex: Seamless Texture Generation as Latent Color Diffusion | [
"Zeqiang Lai",
"Yunfei Zhao",
"Zibo Zhao",
"Xin Yang",
"Xin Huang",
"Jingwei Huang",
"Xiangyu Yue",
"Chunchao Guo"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Lai_NaTex_Seamless_Texture_Generation_as_Latent_Color_Diffusion_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Lai_NaTex_Seamless_Texture_Generation_as_Latent_Color_Diffusion_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Lai_NaTex_Seamless_Texture_CVPR_2026_supplemental.pdf | 2511.16317 | cvf | @InProceedings{Lai_2026_CVPR,
author = {Lai, Zeqiang and Zhao, Yunfei and Zhao, Zibo and Yang, Xin and Huang, Xin and Huang, Jingwei and Yue, Xiangyu and Guo, Chunchao},
title = {NaTex: Seamless Texture Generation as Latent Color Diffusion},
booktitle = {Proceedings of the IEEE/CVF Conference on Comp... | We present NaTex, a native texture generation framework that predicts texture color directly in 3D space. In contrast to previous approaches that rely on baking 2D multi-view images synthesized by geometry-conditioned Multi-View Diffusion models (MVDs), NaTex avoids several inherent limitations of the MVD pipeline. The... | [
0.04196145012974739,
0.004577466286718845,
-0.008971767500042915,
0.06382481008768082,
0.031935036182403564,
0.04475514218211174,
-0.014506859704852104,
0.04214843362569809,
-0.0582234151661396,
-0.09414535760879517,
-0.040883127599954605,
-0.010866576805710793,
-0.02014721743762493,
0.001... |
183 | HoneyBee: Data Recipes for Vision-Language Reasoners | [
"Hritik Bansal",
"Devendra Singh Sachan",
"Kai-Wei Chang",
"Aditya Grover",
"Gargi Ghosh",
"Wen-tau Yih",
"Ramakanth Pasunuru"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Bansal_HoneyBee_Data_Recipes_for_Vision-Language_Reasoners_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Bansal_HoneyBee_Data_Recipes_for_Vision-Language_Reasoners_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Bansal_HoneyBee_Data_Recipes_CVPR_2026_supplemental.pdf | 2510.12225 | cvf | @InProceedings{Bansal_2026_CVPR,
author = {Bansal, Hritik and Sachan, Devendra Singh and Chang, Kai-Wei and Grover, Aditya and Ghosh, Gargi and Yih, Wen-tau and Pasunuru, Ramakanth},
title = {HoneyBee: Data Recipes for Vision-Language Reasoners},
booktitle = {Proceedings of the IEEE/CVF Conference on... | Recent advances in vision-language models (VLMs) have made them highly effective at reasoning tasks. However, the principles underlying the construction of performant VL reasoning training datasets remain poorly understood. In this work, we introduce several data curation approaches and study their impacts on VL reason... | [
0.007617457769811153,
0.00011711387924151495,
-0.005769259762018919,
0.049339037388563156,
0.04670394957065582,
0.01833873987197876,
0.051986902952194214,
0.007688094396144152,
-0.042854588478803635,
-0.017147455364465714,
-0.02816084958612919,
0.035625893622636795,
-0.07649171352386475,
0... |
184 | One-Shot Flow, Any-Time Frame: A Bidirectional Warping Framework for Event-Based Video Frame Interpolation | [
"Linghui Fu",
"Yuhan Liu",
"Hao Chen",
"Zhen Yang",
"Yongjian Deng"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Fu_One-Shot_Flow_Any-Time_Frame_A_Bidirectional_Warping_Framework_for_Event-Based_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Fu_One-Shot_Flow_Any-Time_Frame_A_Bidirectional_Warping_Framework_for_Event-Based_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Fu_One-Shot_Flow_Any-Time_CVPR_2026_supplemental.zip | null | null | @InProceedings{Fu_2026_CVPR,
author = {Fu, Linghui and Liu, Yuhan and Chen, Hao and Yang, Zhen and Deng, Yongjian},
title = {One-Shot Flow, Any-Time Frame: A Bidirectional Warping Framework for Event-Based Video Frame Interpolation},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vis... | Video Frame Interpolation (VFI) is a crucial task in video processing. Flow-based methods, despite their success, are constrained by a fundamental dilemma: forward warping is efficient but prone to artifacts, while backward warping yields higher quality at a significant computational cost, especially for multi-frame in... | [
0.04141052067279816,
-0.018931055441498756,
0.018234893679618835,
0.012824507430195808,
0.025594178587198257,
0.03530188649892807,
0.02088664285838604,
0.021460428833961487,
-0.03330957517027855,
-0.085663340985775,
-0.006063446868211031,
-0.040702760219573975,
-0.0466930978000164,
0.01031... |
185 | Condensed Test-Time Adaptation of VLMs for Action Recognition | [
"Wenxuan Ge",
"Hongyu Qu",
"Rui Yan",
"Guo-Sen Xie",
"Yazhou Yao",
"Xiangbo Shu",
"Jinhui Tang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Ge_Condensed_Test-Time_Adaptation_of_VLMs_for_Action_Recognition_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Ge_Condensed_Test-Time_Adaptation_of_VLMs_for_Action_Recognition_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Ge_Condensed_Test-Time_Adaptation_CVPR_2026_supplemental.pdf | null | null | @InProceedings{Ge_2026_CVPR,
author = {Ge, Wenxuan and Qu, Hongyu and Yan, Rui and Xie, Guo-Sen and Yao, Yazhou and Shu, Xiangbo and Tang, Jinhui},
title = {Condensed Test-Time Adaptation of VLMs for Action Recognition},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Patte... | Test-time adaptation for video understanding, which enables vision-language models (VLMs) to generalize to downstream tasks such as action recognition, has demonstrated substantial value in real-world applications. Existing memory-based methods typically build a visual cache from high-confidence test videos and perform... | [
0.04396803304553032,
-0.019871635362505913,
0.005220259074121714,
0.0361715592443943,
0.060221314430236816,
0.015888549387454987,
0.0412491112947464,
0.025280151516199112,
-0.025500329211354256,
-0.002555791987106204,
-0.01882782205939293,
0.03278791159391403,
-0.06819869577884674,
-0.0100... |
186 | RI-Mamba: Rotation-Invariant Mamba for Robust Text-to-Shape Retrieval | [
"Khanh Nguyen",
"Dasith de Silva Edirimuni",
"Ghulam Mubashar Hassan",
"Ajmal Mian"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Nguyen_RI-Mamba_Rotation-Invariant_Mamba_for_Robust_Text-to-Shape_Retrieval_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Nguyen_RI-Mamba_Rotation-Invariant_Mamba_for_Robust_Text-to-Shape_Retrieval_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Nguyen_RI-Mamba_Rotation-Invariant_Mamba_CVPR_2026_supplemental.pdf | 2602.11673 | cvf | @InProceedings{Nguyen_2026_CVPR,
author = {Nguyen, Khanh and de Silva Edirimuni, Dasith and Hassan, Ghulam Mubashar and Mian, Ajmal},
title = {RI-Mamba: Rotation-Invariant Mamba for Robust Text-to-Shape Retrieval},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Rec... | 3D assets have rapidly expanded in quantity and diversity due to the growing popularity of virtual reality and gaming. As a result, text-to-shape retrieval has become essential in facilitating intuitive search within large repositories. However, existing methods require canonical poses and support few object categories... | [
0.0006417288677766919,
-0.011969263665378094,
-0.003210744122043252,
0.04346639662981033,
0.018631894141435623,
0.03196262568235397,
0.00371938431635499,
0.009897519834339619,
-0.04051448032259941,
-0.0459153838455677,
-0.04036681726574898,
-0.005792359355837107,
-0.09295689314603806,
0.00... |
187 | PointThinker: Point-Incentivized Parallel Thinking for Multimodal Large Language Model | [
"Zhengdong Hu",
"Chao Wang",
"Fengyun Rao",
"Jing LYU",
"Hehe Fan",
"Yi Yang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Hu_PointThinker_Point-Incentivized_Parallel_Thinking_for_Multimodal_Large_Language_Model_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Hu_PointThinker_Point-Incentivized_Parallel_Thinking_for_Multimodal_Large_Language_Model_CVPR_2026_paper.pdf | null | null | null | @InProceedings{Hu_2026_CVPR,
author = {Hu, Zhengdong and Wang, Chao and Rao, Fengyun and LYU, Jing and Fan, Hehe and Yang, Yi},
title = {PointThinker: Point-Incentivized Parallel Thinking for Multimodal Large Language Model},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and ... | This paper explores parallel thinking for Multi-modal Large Language Models (MLLMs), aiming to improve Chain-of-Thought (CoT) through multiple diverse reasoning paths. We guide the model to list multiple visual key points and develop an independent reasoning path for each. Therefore, we term this method PointThinker, w... | [
-0.012897360138595104,
-0.04027010500431061,
0.01271123718470335,
0.03452420234680176,
0.009274538606405258,
0.03115426003932953,
0.015610074624419212,
0.006143263075500727,
-0.0674930065870285,
-0.019685452803969383,
0.003639369970187545,
-0.009470740333199501,
-0.06695577502250671,
0.007... |
188 | Balanced Hierarchical Contrastive Learning with Decoupled Queries for Fine-grained Object Detection in Remote Sensing Images | [
"Jingzhou Chen",
"Dexin Chen",
"Fengchao Xiong",
"Yuntao Qian",
"Liang Xiao"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Chen_Balanced_Hierarchical_Contrastive_Learning_with_Decoupled_Queries_for_Fine-grained_Object_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Chen_Balanced_Hierarchical_Contrastive_Learning_with_Decoupled_Queries_for_Fine-grained_Object_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Chen_Balanced_Hierarchical_Contrastive_CVPR_2026_supplemental.pdf | 2512.24074 | cvf | @InProceedings{Chen_2026_CVPR,
author = {Chen, Jingzhou and Chen, Dexin and Xiong, Fengchao and Qian, Yuntao and Xiao, Liang},
title = {Balanced Hierarchical Contrastive Learning with Decoupled Queries for Fine-grained Object Detection in Remote Sensing Images},
booktitle = {Proceedings of the IEEE/C... | Fine-grained remote sensing datasets often use hierarchical label structures to differentiate objects in a coarse-to-fine manner, with each object annotated across multiple levels. However, embedding this semantic hierarchy into the representation learning space to improve fine-grained detection performance remains cha... | [
0.01521117053925991,
-0.00021913315868005157,
0.022800341248512268,
0.024938639253377914,
0.015471698716282845,
0.018138915300369263,
0.012417026795446873,
-0.013835116289556026,
-0.03681985288858414,
-0.040622930973768234,
-0.03245106711983681,
-0.003214390715584159,
-0.0691021978855133,
... |
189 | Stand-In: A Lightweight and Plug-and-Play Identity Control for Video Generation | [
"Bowen Xue",
"Zheng-Peng Duan",
"Qixin Yan",
"Wenjing Wang",
"Hao Liu",
"Chun-Le Guo",
"Chongyi Li",
"Chen Li",
"Jing Lyu"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Xue_Stand-In_A_Lightweight_and_Plug-and-Play_Identity_Control_for_Video_Generation_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Xue_Stand-In_A_Lightweight_and_Plug-and-Play_Identity_Control_for_Video_Generation_CVPR_2026_paper.pdf | null | 2508.07901 | cvf | @InProceedings{Xue_2026_CVPR,
author = {Xue, Bowen and Duan, Zheng-Peng and Yan, Qixin and Wang, Wenjing and Liu, Hao and Guo, Chun-Le and Li, Chongyi and Li, Chen and Lyu, Jing},
title = {Stand-In: A Lightweight and Plug-and-Play Identity Control for Video Generation},
booktitle = {Proceedings of th... | Generating high-fidelity human videos that match user-specified identities is important yet challenging in the field of generative AI.Existing methods often rely on an excessive number of training parameters and lack compatibility with other AIGC tools.In this paper, we propose Stand-In, a lightweight and plug-and-play... | [
0.014954675920307636,
-0.013554837554693222,
0.0035962958354502916,
0.0419524684548378,
0.04730929061770439,
0.011458058841526508,
0.007256526034325361,
-0.01525610126554966,
-0.0328892357647419,
-0.04302351549267769,
-0.027034670114517212,
-0.011986259371042252,
-0.06599687784910202,
-0.0... |
190 | SEASON: Mitigating Temporal Hallucination in Video Large Language Models via Self-Diagnostic Contrastive Decoding | [
"Chang-Hsun Wu",
"Kai-Po Chang",
"Yu-Yang Sheng",
"Hung-Kai Chung",
"Kuei-Chun Wang",
"Yu-Chiang Frank Wang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Wu_SEASON_Mitigating_Temporal_Hallucination_in_Video_Large_Language_Models_via_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Wu_SEASON_Mitigating_Temporal_Hallucination_in_Video_Large_Language_Models_via_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Wu_SEASON_Mitigating_Temporal_CVPR_2026_supplemental.pdf | 2512.04643 | cvf | @InProceedings{Wu_2026_CVPR,
author = {Wu, Chang-Hsun and Chang, Kai-Po and Sheng, Yu-Yang and Chung, Hung-Kai and Wang, Kuei-Chun and Wang, Yu-Chiang Frank},
title = {SEASON: Mitigating Temporal Hallucination in Video Large Language Models via Self-Diagnostic Contrastive Decoding},
booktitle = {Proc... | Video Large Language Models (VideoLLMs) have shown remarkable progress in video understanding. However, these models still struggle to effectively perceive and exploit rich temporal information in videos when responding to user queries. Therefore, they often generate descriptions of events that are temporal inconsisten... | [
0.02390669286251068,
0.015285640023648739,
0.009268372319638729,
0.03680431842803955,
0.030724022537469864,
0.006532707251608372,
0.06574749946594238,
0.02871161885559559,
-0.05597789213061333,
-0.024028630927205086,
-0.029668651521205902,
0.008756650611758232,
-0.04577309638261795,
0.0140... |
191 | VecGlypher: Unified Vector Glyph Generation with Language Models | [
"Xiaoke Huang",
"Bhavul Gauri",
"Kam Woh Ng",
"Tony Ng",
"Mengmeng Xu",
"Zhiheng Liu",
"Weiming Ren",
"Zhaochong An",
"Zijian Zhou",
"Haonan Qiu",
"Yuyin Zhou",
"Sen He",
"Ziheng Wang",
"Tao Xiang",
"Xiao Han"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Huang_VecGlypher_Unified_Vector_Glyph_Generation_with_Language_Models_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Huang_VecGlypher_Unified_Vector_Glyph_Generation_with_Language_Models_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Huang_VecGlypher_Unified_Vector_CVPR_2026_supplemental.pdf | 2602.21461 | cvf | @InProceedings{Huang_2026_CVPR,
author = {Huang, Xiaoke and Gauri, Bhavul and Ng, Kam Woh and Ng, Tony and Xu, Mengmeng and Liu, Zhiheng and Ren, Weiming and An, Zhaochong and Zhou, Zijian and Qiu, Haonan and Zhou, Yuyin and He, Sen and Wang, Ziheng and Xiang, Tao and Han, Xiao},
title = {VecGlypher: Uni... | Vector glyphs are the atomic units of digital typography, yet most learning-based pipelines still depend on carefully curated exemplar sheets and raster-to-vector postprocessing, which limits accessibility and editability. We introduce VecGlypher, a single multimodal language model that generates high-fidelity vector g... | [
0.021777179092168808,
-0.011453804560005665,
0.02308911457657814,
0.06592780351638794,
0.02214216999709606,
0.050847843289375305,
0.026675349101424217,
0.0559704452753067,
-0.03250470012426376,
-0.061098337173461914,
-0.02741730399429798,
-0.008131703361868858,
-0.06590188294649124,
0.0058... |
192 | MMSD3.0: A Multi-Image Benchmark for Real-World Multimodal Sarcasm Detection | [
"Haochen Zhao",
"Yuyao Kong",
"Yongxiu Xu",
"Gaopeng Gou",
"Hongbo Xu",
"Yubin Wang",
"Haoliang Zhang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Zhao_MMSD3.0_A_Multi-Image_Benchmark_for_Real-World_Multimodal_Sarcasm_Detection_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Zhao_MMSD3.0_A_Multi-Image_Benchmark_for_Real-World_Multimodal_Sarcasm_Detection_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zhao_MMSD3.0_A_Multi-Image_CVPR_2026_supplemental.pdf | 2510.23299 | title_snapshot | @InProceedings{Zhao_2026_CVPR,
author = {Zhao, Haochen and Kong, Yuyao and Xu, Yongxiu and Gou, Gaopeng and Xu, Hongbo and Wang, Yubin and Zhang, Haoliang},
title = {MMSD3.0: A Multi-Image Benchmark for Real-World Multimodal Sarcasm Detection},
booktitle = {Proceedings of the IEEE/CVF Conference on C... | Despite progress in multimodal sarcasm detection, existing datasets and methods predominantly focus on single-image scenarios, overlooking potential semantic and affective relations across multiple images. This leaves a gap in modeling cases where sarcasm is triggered by multi-image cues in real-world settings. To brid... | [
0.007429559249430895,
-0.014426358044147491,
0.007139763794839382,
0.06292940676212311,
0.02163437008857727,
0.019390400499105453,
0.033418986946344376,
0.019025979563593864,
-0.036732837557792664,
-0.011850416660308838,
-0.014388175681233406,
0.04061513766646385,
-0.04371476545929909,
-0.... |
193 | Fusion in Your Way: Aligning Image Fusion with Heterogeneous Demands via Direct Preference Optimization | [
"Weijian Su",
"Songqian Zhang",
"Yuqi Han",
"Jian Zhuang",
"Yongdong Huang",
"Qiang Zhang"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Su_Fusion_in_Your_Way_Aligning_Image_Fusion_with_Heterogeneous_Demands_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Su_Fusion_in_Your_Way_Aligning_Image_Fusion_with_Heterogeneous_Demands_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Su_Fusion_in_Your_CVPR_2026_supplemental.pdf | 2605.06049 | cvf | @InProceedings{Su_2026_CVPR,
author = {Su, Weijian and Zhang, Songqian and Han, Yuqi and Zhuang, Jian and Huang, Yongdong and Zhang, Qiang},
title = {Fusion in Your Way: Aligning Image Fusion with Heterogeneous Demands via Direct Preference Optimization},
booktitle = {Proceedings of the IEEE/CVF Conf... | As a key technique in multi-modal processing, infrared and visible image fusion (IVIF) plays a crucial role in integrating complementary spectral information for visual enhancement and downstream vision tasks. Despite remarkable progress, existing methods struggle to flexibly accommodate heterogeneous demands. Achievin... | [
-0.014610781334340572,
0.005949791520833969,
0.013563449494540691,
0.04117727652192116,
0.028149785473942757,
0.04591251537203789,
0.007265447173267603,
0.01739366538822651,
-0.017860669642686844,
-0.07434911280870438,
-0.019151102751493454,
0.014932921156287193,
-0.07924976944923401,
-0.0... |
194 | A Training-Free Style-Personalization via SVD-Based Feature Decomposition | [
"Kyoungmin Lee",
"Jihun Park",
"Jongmin Gim",
"Wonhyeok Choi",
"Kyumin Hwang",
"Jaeyeul Kim",
"Sunghoon Im"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Lee_A_Training-Free_Style-Personalization_via_SVD-Based_Feature_Decomposition_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Lee_A_Training-Free_Style-Personalization_via_SVD-Based_Feature_Decomposition_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Lee_A_Training-Free_Style-Personalization_CVPR_2026_supplemental.pdf | 2507.04482 | cvf | @InProceedings{Lee_2026_CVPR,
author = {Lee, Kyoungmin and Park, Jihun and Gim, Jongmin and Choi, Wonhyeok and Hwang, Kyumin and Kim, Jaeyeul and Im, Sunghoon},
title = {A Training-Free Style-Personalization via SVD-Based Feature Decomposition},
booktitle = {Proceedings of the IEEE/CVF Conference on ... | We present a training-free framework for style-personalized image generation that operates during inference using a scale-wise autoregressive model. Our method generates a stylized image guided by a single reference style while preserving semantic consistency and mitigating content leakage. Through a detailed step-wise... | [
0.03980289399623871,
-0.01581273041665554,
0.0285387821495533,
0.04825051128864288,
0.05431191623210907,
0.045094817876815796,
0.02446948178112507,
-0.0072767469100654125,
-0.03140078857541084,
-0.06968661397695541,
-0.052016016095876694,
0.008192909881472588,
-0.06928596645593643,
0.02056... |
195 | Beyond the Ground Truth: Enhanced Supervision for Image Restoration | [
"Donghun Ryou",
"Inju Ha",
"Sanghyeok Chu",
"Bohyung Han"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Ryou_Beyond_the_Ground_Truth_Enhanced_Supervision_for_Image_Restoration_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Ryou_Beyond_the_Ground_Truth_Enhanced_Supervision_for_Image_Restoration_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Ryou_Beyond_the_Ground_CVPR_2026_supplemental.pdf | 2512.03932 | cvf | @InProceedings{Ryou_2026_CVPR,
author = {Ryou, Donghun and Ha, Inju and Chu, Sanghyeok and Han, Bohyung},
title = {Beyond the Ground Truth: Enhanced Supervision for Image Restoration},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month ... | Deep learning-based image restoration has achieved significant success. However, when addressing real-world degradations, model performance is limited by the quality of ground-truth images in datasets due to practical constraints in data acquisition. To address this limitation, we propose a novel framework that enhance... | [
0.018361881375312805,
-0.004295933526009321,
-0.003257715841755271,
0.05465938150882721,
0.061355773359537125,
0.001572348177433014,
0.010309058241546154,
0.00167843047529459,
-0.057517945766448975,
-0.07415307313203812,
-0.04122684523463249,
0.020717110484838486,
-0.060793135315179825,
0.... |
196 | SineProject: Machine Unlearning for Stable Vision-Language Alignment | [
"Arpit Garg",
"Hemanth Saratchandran",
"Simon Lucey"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Garg_SineProject_Machine_Unlearning_for_Stable_Vision-Language_Alignment_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Garg_SineProject_Machine_Unlearning_for_Stable_Vision-Language_Alignment_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Garg_SineProject_Machine_Unlearning_CVPR_2026_supplemental.pdf | 2511.18444 | cvf | @InProceedings{Garg_2026_CVPR,
author = {Garg, Arpit and Saratchandran, Hemanth and Lucey, Simon},
title = {SineProject: Machine Unlearning for Stable Vision-Language Alignment},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {J... | Multimodal Large Language Models (MLLMs) increasingly need to forget specific knowledge, such as unsafe or private information, without full retraining. However, existing unlearning methods often disrupt vision-language alignment, causing models to reject both harmful and benign queries simultaneously. We trace this fa... | [
0.019371407106518745,
-0.003969870042055845,
-0.006883199326694012,
0.027142755687236786,
0.015499402768909931,
0.005544072948396206,
0.047848571091890335,
0.04414116591215134,
-0.044300030916929245,
-0.012822170741856098,
-0.022192956879734993,
0.040061306208372116,
-0.08438760042190552,
... |
197 | Unlocking Token Rewards via Training-Free Reward Attribution | [
"Sitong Wu",
"Haoru Tan",
"Bin Xia",
"Xichen Zhang",
"Jingyao Li",
"Shaofeng Zhang",
"Xiaojuan Qi",
"Bei Yu",
"Jiaya Jia"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Wu_Unlocking_Token_Rewards_via_Training-Free_Reward_Attribution_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Wu_Unlocking_Token_Rewards_via_Training-Free_Reward_Attribution_CVPR_2026_paper.pdf | null | null | null | @InProceedings{Wu_2026_CVPR,
author = {Wu, Sitong and Tan, Haoru and Xia, Bin and Zhang, Xichen and Li, Jingyao and Zhang, Shaofeng and Qi, Xiaojuan and Yu, Bei and Jia, Jiaya},
title = {Unlocking Token Rewards via Training-Free Reward Attribution},
booktitle = {Proceedings of the IEEE/CVF Conference... | In this paper, we propose an extremely efficient, training-free method to extract token-level reward signals directly from an existing deep reward model. Our core idea is to attribute the overall process reward to individual tokens by estimating each token's influence. This influence is defined as the change in the fin... | [
-0.0005707746022380888,
-0.03146492689847946,
0.014525562524795532,
0.038756113499403,
0.02923852577805519,
0.019632190465927124,
-0.006940467748790979,
-0.003956885077059269,
-0.016572291031479836,
-0.023287246003746986,
-0.03249192237854004,
0.02526121214032173,
-0.046536121517419815,
-0... |
198 | Delving Aleatoric Uncertainty in Medical Image Segmentation via Vision Foundation Models | [
"Ruiyang Li",
"Fang Liu",
"Licheng Jiao",
"Xinglin Xie",
"Jiayao Hao",
"Shuo Li",
"Xu Liu",
"Jingyi Yang",
"Lingling Li",
"Puhua Chen",
"Wenping Ma"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Li_Delving_Aleatoric_Uncertainty_in_Medical_Image_Segmentation_via_Vision_Foundation_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Li_Delving_Aleatoric_Uncertainty_in_Medical_Image_Segmentation_via_Vision_Foundation_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Li_Delving_Aleatoric_Uncertainty_CVPR_2026_supplemental.pdf | 2604.10963 | cvf | @InProceedings{Li_2026_CVPR,
author = {Li, Ruiyang and Liu, Fang and Jiao, Licheng and Xie, Xinglin and Hao, Jiayao and Li, Shuo and Liu, Xu and Yang, Jingyi and Li, Lingling and Chen, Puhua and Ma, Wenping},
title = {Delving Aleatoric Uncertainty in Medical Image Segmentation via Vision Foundation Model... | Medical image segmentation supports clinical workflows by precisely delineating anatomical structures and lesions. However, medical image datasets medical image datasets suffer from acquisition noise and annotation ambiguity, causing pervasive data uncertainty that substantially undermines model robustness. Existing re... | [
0.009342311881482601,
-0.007416766602545977,
-0.020934179425239563,
0.042215652763843536,
0.042693041265010834,
0.03661460429430008,
0.017672846093773842,
0.031414661556482315,
-0.042396429926157,
-0.09202810376882553,
-0.014854657463729382,
-0.013192703016102314,
-0.05031586065888405,
0.0... |
199 | Language Models Can Explain Visual Features via Steering | [
"Javier Ferrando",
"Enrique Lopez-Cuena",
"Pablo Agustin Martin-Torres",
"Daniel Hinjos",
"Anna Arias-Duart",
"Dario Garcia-Gasulla"
] | https://openaccess.thecvf.com/content/CVPR2026/html/Ferrando_Language_Models_Can_Explain_Visual_Features_via_Steering_CVPR_2026_paper.html | https://openaccess.thecvf.com/content/CVPR2026/papers/Ferrando_Language_Models_Can_Explain_Visual_Features_via_Steering_CVPR_2026_paper.pdf | https://openaccess.thecvf.com/content/CVPR2026/supplemental/Ferrando_Language_Models_Can_CVPR_2026_supplemental.pdf | 2603.22593 | title_snapshot | @InProceedings{Ferrando_2026_CVPR,
author = {Ferrando, Javier and Lopez-Cuena, Enrique and Martin-Torres, Pablo Agustin and Hinjos, Daniel and Arias-Duart, Anna and Garcia-Gasulla, Dario},
title = {Language Models Can Explain Visual Features via Steering},
booktitle = {Proceedings of the IEEE/CVF Con... | Sparse Autoencoders uncover thousands of features in vision models, yet explaining these features without requiring human intervention remains an open challenge. While previous work has proposed generating correlation-based explanations based on top activating input examples, we present a fundamentally different altern... | [
-0.00047523045213893056,
-0.0025062267668545246,
0.006620653439313173,
0.04590470343828201,
0.015194134786725044,
0.04531250149011612,
0.035962946712970734,
0.024572648108005524,
-0.0176070649176836,
-0.02379918098449707,
-0.04924708604812622,
0.00900929607450962,
-0.05284512788057327,
0.0... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.