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
4,000
PhysGM: Large Physical Gaussian Model for Feed-Forward 4D Synthesis
[ "Chunji Lv", "Zequn Chen", "Donglin Di", "Weinan Zhang", "Hao Li", "Chen Wei", "Yinjie Lei", "Changsheng Li" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Lv_PhysGM_Large_Physical_Gaussian_Model_for_Feed-Forward_4D_Synthesis_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Lv_PhysGM_Large_Physical_Gaussian_Model_for_Feed-Forward_4D_Synthesis_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Lv_PhysGM_Large_Physical_CVPR_2026_supplemental.zip
2508.13911
cvf
@InProceedings{Lv_2026_CVPR, author = {Lv, Chunji and Chen, Zequn and Di, Donglin and Zhang, Weinan and Li, Hao and Wei, Chen and Lei, Yinjie and Li, Changsheng}, title = {PhysGM: Large Physical Gaussian Model for Feed-Forward 4D Synthesis}, booktitle = {Proceedings of the IEEE/CVF Conference on Comp...
Despite advances in physics-based 3D motion synthesis, current methods face key limitations: reliance on pre-reconstructed 3D Gaussian Splatting (3DGS) built from dense multi-view images with time-consuming per-scene optimization; physics integration via either inflexible, hand-specified attributes or unstable, optimiz...
[ -0.00046551262494176626, 0.0110356155782938, 0.018777187913656235, 0.055937208235263824, 0.01438367459923029, 0.023347236216068268, 0.0010581482201814651, 0.013506033457815647, -0.045261017978191376, -0.046035654842853546, -0.006186131853610277, -0.022175535559654236, -0.059300702065229416, ...
4,001
EmoStyle: Emotion-Driven Image Stylization
[ "Jingyuan Yang", "Zihuan Bai", "Hui Huang" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Yang_EmoStyle_Emotion-Driven_Image_Stylization_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Yang_EmoStyle_Emotion-Driven_Image_Stylization_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Yang_EmoStyle_Emotion-Driven_Image_CVPR_2026_supplemental.pdf
2512.05478
cvf
@InProceedings{Yang_2026_CVPR, author = {Yang, Jingyuan and Bai, Zihuan and Huang, Hui}, title = {EmoStyle: Emotion-Driven Image Stylization}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = {June}, year = {2026}, pa...
Art has long been a profound medium for expressing emotions.While existing image stylization methods effectively transform visual appearance, they often overlook the emotional impact carried by styles.To bridge this gap, we introduce Affective Image Stylization (AIS), a task that applies artistic styles to evoke specif...
[ 0.02367609553039074, -0.042606476694345474, -0.0064941407181322575, 0.03675982728600502, 0.008484621532261372, 0.05023909732699394, 0.014718063175678253, 0.007776016369462013, -0.021973559632897377, -0.0672960877418518, -0.04065121337771416, -0.00038517676875926554, -0.03729952126741409, -...
4,002
AcTTA: Rethinking Test-Time Adaptation via Dynamic Activation
[ "Hyeongyu Kim", "Geonhui Han", "Dosik Hwang" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Kim_AcTTA_Rethinking_Test-Time_Adaptation_via_Dynamic_Activation_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Kim_AcTTA_Rethinking_Test-Time_Adaptation_via_Dynamic_Activation_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Kim_AcTTA_Rethinking_Test-Time_CVPR_2026_supplemental.pdf
2603.26096
cvf
@InProceedings{Kim_2026_CVPR, author = {Kim, Hyeongyu and Han, Geonhui and Hwang, Dosik}, title = {AcTTA: Rethinking Test-Time Adaptation via Dynamic Activation}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = {June}, year ...
Test-time adaptation (TTA) aims to mitigate performance degradation under distribution shifts by updating model parameters during inference. Existing approaches have primarily framed adaptation around affine modulation, focusing on recalibrating normalization layers. This perspective, while effective, overlooks another...
[ 0.0063703348860144615, -0.0317530557513237, 0.007907318882644176, 0.029875457286834717, 0.03696630895137787, 0.03341345116496086, 0.0423443540930748, -0.02096550166606903, -0.0006707413704134524, -0.014207372441887856, 0.021430427208542824, 0.0034084690269082785, -0.05483512580394745, -0.0...
4,003
Prospective Dynamic 3D MRI Reconstruction via Latent-Space Motion Tracking from Single Measurement
[ "Lixuan Chen", "Zhongnan Liu", "Jesse Hamilton", "James M. Balter", "Jeong Joon Park", "Liyue Shen" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Chen_Prospective_Dynamic_3D_MRI_Reconstruction_via_Latent-Space_Motion_Tracking_from_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Chen_Prospective_Dynamic_3D_MRI_Reconstruction_via_Latent-Space_Motion_Tracking_from_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Chen_Prospective_Dynamic_3D_CVPR_2026_supplemental.zip
null
null
@InProceedings{Chen_2026_CVPR, author = {Chen, Lixuan and Liu, Zhongnan and Hamilton, Jesse and Balter, James M. and Park, Jeong Joon and Shen, Liyue}, title = {Prospective Dynamic 3D MRI Reconstruction via Latent-Space Motion Tracking from Single Measurement}, booktitle = {Proceedings of the IEEE/CV...
Prospective reconstruction is crucial in many clinical applications such as MRI-guided radiotherapy, which demands accurate image reconstruction and fast motion estimation from currently acquired measurements. However, prospective reconstruction remains challenging due to ultra-sparse sampling and stringent latency req...
[ -0.042310863733291626, -0.02571694552898407, 0.008999221958220005, 0.017922736704349518, 0.042008768767118454, 0.03373401612043381, 0.02474590577185154, 0.010022860020399094, -0.043820492923259735, -0.08036945760250092, 0.005178058985620737, -0.03752165660262108, -0.008194202557206154, 0.0...
4,004
STARFlow-V: End-to-End Video Generative Modeling with Autoregressive Normalizing Flows
[ "Jiatao Gu", "Ying Shen", "Tianrong Chen", "Laurent Dinh", "Yuyang Wang", "Miguel Angel Bautista", "David Berthelot", "Josh Susskind", "Shuangfei Zhai" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Gu_STARFlow-V_End-to-End_Video_Generative_Modeling_with_Autoregressive_Normalizing_Flows_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Gu_STARFlow-V_End-to-End_Video_Generative_Modeling_with_Autoregressive_Normalizing_Flows_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Gu_STARFlow-V_End-to-End_Video_CVPR_2026_supplemental.zip
2511.20462
title_judge
@InProceedings{Gu_2026_CVPR, author = {Gu, Jiatao and Shen, Ying and Chen, Tianrong and Dinh, Laurent and Wang, Yuyang and Bautista, Miguel Angel and Berthelot, David and Susskind, Josh and Zhai, Shuangfei}, title = {STARFlow-V: End-to-End Video Generative Modeling with Autoregressive Normalizing Flows},...
Normalizing flows (NFs) are end-to-end likelihood-based generative models for continuous data, and have recently regained attention with encouraging progress on image generation. Yet in the video generation domain, where spatiotemporal complexity and computational cost are substantially higher, state-of-the-art systems...
[ 0.03395254909992218, -0.0361916720867157, 0.029334837570786476, 0.053931768983602524, 0.041815415024757385, 0.05247717350721359, 0.02924550138413906, -0.007594279479235411, -0.01480829156935215, -0.07178907841444016, -0.005210842005908489, -0.046146176755428314, -0.0424385741353035, 0.0273...
4,005
High-Fidelity Virtual Try-On beyond Paired Data Scarcity via Diffusion-based Cycle-Consistent Learning
[ "Jia Wu", "Yijing Dai", "Tingfeng Cao", "Meiling Wu", "Tao Luo", "Jian Dong Zhang", "Guangming Lu", "Xiaoyi Zeng" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Wu_High-Fidelity_Virtual_Try-On_beyond_Paired_Data_Scarcity_via_Diffusion-based_Cycle-Consistent_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Wu_High-Fidelity_Virtual_Try-On_beyond_Paired_Data_Scarcity_via_Diffusion-based_Cycle-Consistent_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Wu_High-Fidelity_Virtual_Try-On_CVPR_2026_supplemental.pdf
null
null
@InProceedings{Wu_2026_CVPR, author = {Wu, Jia and Dai, Yijing and Cao, Tingfeng and Wu, Meiling and Luo, Tao and Zhang, Jian Dong and Lu, Guangming and Zeng, Xiaoyi}, title = {High-Fidelity Virtual Try-On beyond Paired Data Scarcity via Diffusion-based Cycle-Consistent Learning}, booktitle = {Procee...
Diffusion-based virtual try-on methods rely on vast high-quality garment-person pairs, which are scarce in practice due to the high cost of data collection and preprocessing, limiting their performance in real-world scenarios.To overcome this bottleneck, we propose Cycle-Consistent Virtual Try-On (CCVTON), a diffusion-...
[ 0.05041873827576637, -0.021706899628043175, -0.029853178188204765, 0.040856968611478806, 0.05826275423169136, 0.0407242514193058, 0.03776387870311737, 0.0120431799441576, -0.0023779310286045074, -0.05979949235916138, -0.033831339329481125, -0.020742878317832947, -0.059053488075733185, -0.0...
4,006
Self-guided Semantic Inspection for Zero-Shot Composed Image Retrieval
[ "Jingjing Zhang", "Lei Zhang", "Zheren Fu", "Bo Hu", "Zhendong Mao" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Zhang_Self-guided_Semantic_Inspection_for_Zero-Shot_Composed_Image_Retrieval_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Zhang_Self-guided_Semantic_Inspection_for_Zero-Shot_Composed_Image_Retrieval_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zhang_Self-guided_Semantic_Inspection_CVPR_2026_supplemental.pdf
null
null
@InProceedings{Zhang_2026_CVPR, author = {Zhang, Jingjing and Zhang, Lei and Fu, Zheren and Hu, Bo and Mao, Zhendong}, title = {Self-guided Semantic Inspection for Zero-Shot Composed Image Retrieval}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR...
Zero-Shot Composed Image Retrieval (ZS-CIR) aims to retrieve target images using a composed query of a reference image and a textual modification, without relying on triplet-based supervision. As the two inputs describe related but semantically unaligned information, the key challenge lies in interpreting their cross-m...
[ 0.018864423036575317, -0.023111335933208466, -0.028734011575579643, 0.05686849355697632, 0.0433996319770813, 0.0020786349195986986, 0.020116474479436874, 0.030505970120429993, -0.028998641297221184, -0.030056536197662354, -0.05518858879804611, 0.018123526126146317, -0.05191362649202347, 0....
4,007
Vision-Language Attribute Disentanglement and Reinforcement for Lifelong Person Re-Identification
[ "Kunlun Xu", "Haotong Cheng", "Jiangmeng Li", "Xu Zou", "Jiahuan Zhou" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Xu_Vision-Language_Attribute_Disentanglement_and_Reinforcement_for_Lifelong_Person_Re-Identification_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Xu_Vision-Language_Attribute_Disentanglement_and_Reinforcement_for_Lifelong_Person_Re-Identification_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Xu_Vision-Language_Attribute_Disentanglement_CVPR_2026_supplemental.pdf
2603.19678
cvf
@InProceedings{Xu_2026_CVPR, author = {Xu, Kunlun and Cheng, Haotong and Li, Jiangmeng and Zou, Xu and Zhou, Jiahuan}, title = {Vision-Language Attribute Disentanglement and Reinforcement for Lifelong Person Re-Identification}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision an...
Lifelong person re-identification (LReID) aims to learn from varying domains to obtain a unified person retrieval model. Existing LReID approaches typically focus on learning from scratch or a visual classification-pretrained model, while the Vision-Language Model (VLM) has shown generalizable knowledge in a variety of...
[ 0.01654079556465149, -0.010013153776526451, 0.048255737870931625, 0.05659596249461174, 0.031887657940387726, -0.022719256579875946, 0.028264766559004784, -0.004288441967219114, -0.005187630653381348, -0.021349657326936722, -0.051052093505859375, 0.029748031869530678, -0.08765900135040283, ...
4,008
SAGA: Source Attribution of Generative AI Videos
[ "Rohit Kundu", "Vishal Mohanty", "Hao Xiong", "Shan Jia", "Athula Balachandran", "Amit K. Roy-Chowdhury" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Kundu_SAGA_Source_Attribution_of_Generative_AI_Videos_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Kundu_SAGA_Source_Attribution_of_Generative_AI_Videos_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Kundu_SAGA_Source_Attribution_CVPR_2026_supplemental.pdf
2511.12834
cvf
@InProceedings{Kundu_2026_CVPR, author = {Kundu, Rohit and Mohanty, Vishal and Xiong, Hao and Jia, Shan and Balachandran, Athula and Roy-Chowdhury, Amit K.}, title = {SAGA: Source Attribution of Generative AI Videos}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern ...
The proliferation of generative AI has led to hyper-realistic synthetic videos, escalating misuse risks and outstripping binary real/fake detectors. We introduce \texttt SAGA (\underline S ource \underline A ttribution of \underline G enerative \underline A I videos), the first comprehensive framework to address the u...
[ 0.05882926285266876, -0.02472127601504326, -0.003124753013253212, 0.08526964485645294, 0.015568507835268974, 0.015634644776582718, 0.028116075322031975, 0.004486437886953354, 0.0080866739153862, -0.03154350444674492, -0.026798810809850693, 0.019846439361572266, -0.048508405685424805, 0.006...
4,009
Dance Across Shifts: Forward-Facilitation Continual Test-Time Adaptation through Dynamic Style Bridging
[ "Zhilin Zhu", "Yabin Wang", "Zhiheng Ma", "Yaguang Song", "Yaowei Wang", "Xiaopeng Hong" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Zhu_Dance_Across_Shifts_Forward-Facilitation_Continual_Test-Time_Adaptation_through_Dynamic_Style_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Zhu_Dance_Across_Shifts_Forward-Facilitation_Continual_Test-Time_Adaptation_through_Dynamic_Style_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zhu_Dance_Across_Shifts_CVPR_2026_supplemental.pdf
2605.18608
cvf
@InProceedings{Zhu_2026_CVPR, author = {Zhu, Zhilin and Wang, Yabin and Ma, Zhiheng and Song, Yaguang and Wang, Yaowei and Hong, Xiaopeng}, title = {Dance Across Shifts: Forward-Facilitation Continual Test-Time Adaptation through Dynamic Style Bridging}, booktitle = {Proceedings of the IEEE/CVF Confe...
Continual Test-Time Adaptation (CTTA) aims to empower perception systems to handle dynamic distribution shifts encountered after deployment. Existing methods predominantly follow a backward-alignment paradigm, which rigidly aligns incoming data with supervisory surrogates derived from the source domain. Consequently, t...
[ 0.029100414365530014, -0.018305383622646332, -0.024041900411248207, 0.031874269247055054, 0.04514586925506592, 0.011316027492284775, 0.045303720980882645, 0.019396938383579254, -0.029052849858999252, -0.03707963228225708, -0.03520336374640465, 0.007910779677331448, -0.08273980021476746, -0...
4,010
Enhancing Continual Learning of Vision-Language Models via Dynamic Prefix Weighting
[ "Hyeonseo Jang", "Hyuk Kwon", "Kibok Lee" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Jang_Enhancing_Continual_Learning_of_Vision-Language_Models_via_Dynamic_Prefix_Weighting_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Jang_Enhancing_Continual_Learning_of_Vision-Language_Models_via_Dynamic_Prefix_Weighting_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Jang_Enhancing_Continual_Learning_CVPR_2026_supplemental.pdf
2604.18075
cvf
@InProceedings{Jang_2026_CVPR, author = {Jang, Hyeonseo and Kwon, Hyuk and Lee, Kibok}, title = {Enhancing Continual Learning of Vision-Language Models via Dynamic Prefix Weighting}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month ...
We investigate recently introduced domain-class incremental learning scenarios for vision-language models (VLMs). Recent works address this challenge using parameter-efficient methods, such as prefix-tuning or adapters, which facilitate model adaptation to downstream tasks by incorporating task-specific information int...
[ 0.0013780525187030435, -0.008141711354255676, 0.003621649229899049, 0.03273420035839081, 0.02266683056950569, 0.047097347676754, 0.02647976577281952, -0.007254740688949823, -0.03622641786932945, -0.04206441715359688, -0.0234784334897995, 0.030297761783003807, -0.08546169847249985, 0.009265...
4,011
Temporal Inversion for Learning Interval Change in Chest X-Rays
[ "Hanbin Ko", "Kyeongmin Jeon", "Doowoong Choi", "Chang Min Park" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Ko_Temporal_Inversion_for_Learning_Interval_Change_in_Chest_X-Rays_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Ko_Temporal_Inversion_for_Learning_Interval_Change_in_Chest_X-Rays_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Ko_Temporal_Inversion_for_CVPR_2026_supplemental.pdf
2604.04563
cvf
@InProceedings{Ko_2026_CVPR, author = {Ko, Hanbin and Jeon, Kyeongmin and Choi, Doowoong and Park, Chang Min}, title = {Temporal Inversion for Learning Interval Change in Chest X-Rays}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month ...
Recent advances in vision-language pretraining have enabled strong medical foundation models, yet most analyze radiographs in isolation, overlooking the key clinical task of comparing prior and current images to assess interval change. For chest radiographs (CXRs), capturing interval change is essential, as radiologist...
[ 0.022272193804383278, -0.004657006356865168, -0.051092322915792465, 0.029829030856490135, 0.03900928050279617, 0.008745736442506313, 0.025292687118053436, 0.017282413318753242, -0.021584689617156982, -0.03922034054994583, -0.01856103539466858, -0.00573108671233058, -0.03018965944647789, 0....
4,012
PixARMesh: Autoregressive Mesh-Native Single-View Scene Reconstruction
[ "Xiang Zhang", "Sohyun Yoo", "Hongrui Wu", "Chuan Li", "Jianwen Xie", "Zhuowen Tu" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Zhang_PixARMesh_Autoregressive_Mesh-Native_Single-View_Scene_Reconstruction_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Zhang_PixARMesh_Autoregressive_Mesh-Native_Single-View_Scene_Reconstruction_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zhang_PixARMesh_Autoregressive_Mesh-Native_CVPR_2026_supplemental.pdf
2603.05888
cvf
@InProceedings{Zhang_2026_CVPR, author = {Zhang, Xiang and Yoo, Sohyun and Wu, Hongrui and Li, Chuan and Xie, Jianwen and Tu, Zhuowen}, title = {PixARMesh: Autoregressive Mesh-Native Single-View Scene Reconstruction}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern ...
We introduce PixARMesh, a method to autoregressively reconstruct complete 3D indoor scene meshes directly from a single RGB image. Unlike prior methods that rely on implicit signed distance fields and post-hoc layout optimization, PixARMesh jointly predicts object layout and geometry within a unified model, producing c...
[ 0.014586935751140118, 0.0037292588967829943, -0.012662877328693867, 0.020495085045695305, 0.0354955717921257, 0.07350166887044907, 0.030149662867188454, 0.020703956484794617, -0.05583423748612404, -0.08214698731899261, -0.012781227007508278, -0.049118250608444214, -0.08079183846712112, -0....
4,013
PS-SR: Pseudo-Single-Step Video Super-Resolution via Speculative Diffusion
[ "Aiqiu Wu", "Zhaofan Qiu", "Ting Yao", "Tao Mei" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Wu_PS-SR_Pseudo-Single-Step_Video_Super-Resolution_via_Speculative_Diffusion_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Wu_PS-SR_Pseudo-Single-Step_Video_Super-Resolution_via_Speculative_Diffusion_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Wu_PS-SR_Pseudo-Single-Step_Video_CVPR_2026_supplemental.pdf
null
null
@InProceedings{Wu_2026_CVPR, author = {Wu, Aiqiu and Qiu, Zhaofan and Yao, Ting and Mei, Tao}, title = {PS-SR: Pseudo-Single-Step Video Super-Resolution via Speculative Diffusion}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = ...
Video Super-Resolution (VSR) fundamentally struggles with a critical trade-off: single-step models offer unmatched efficiency but often lack the high-frequency detail, creativity, and visual quality of their multi-step diffusion counterparts, which are computationally prohibitive for practical use. In this paper, we pr...
[ 0.005292424466460943, -0.004081288818269968, 0.019000446423888206, 0.052591923624277115, 0.05043923109769821, 0.02477157488465309, 0.016686268150806427, -0.017317969352006912, -0.03587781637907028, -0.056099556386470795, 0.0037284493446350098, -0.033942047506570816, -0.04261220246553421, 0...
4,014
ARES: Unifying Asymmetric RGB-Event Stereo for Probabilistic Scene Flow Estimation
[ "Jie Long Lee", "Gim Hee Lee" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Lee_ARES_Unifying_Asymmetric_RGB-Event_Stereo_for_Probabilistic_Scene_Flow_Estimation_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Lee_ARES_Unifying_Asymmetric_RGB-Event_Stereo_for_Probabilistic_Scene_Flow_Estimation_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Lee_ARES_Unifying_Asymmetric_CVPR_2026_supplemental.pdf
null
null
@InProceedings{Lee_2026_CVPR, author = {Lee, Jie Long and Lee, Gim Hee}, title = {ARES: Unifying Asymmetric RGB-Event Stereo for Probabilistic Scene Flow Estimation}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = {June}, ye...
Estimating dense three dimensional motion in dynamic high speed scenes remains challenging due to motion blur, illumination variation, and the limited temporal resolution of conventional cameras. We introduce ARES, a unified framework for Asymmetric RGB-Event Stereo that addresses these issues through a hybrid setup wh...
[ 0.031684182584285736, 0.004640430677682161, -0.0034683484118431807, 0.01693187654018402, 0.012974233366549015, 0.03584383428096771, 0.016745803877711296, 0.05053842440247536, -0.05444353073835373, -0.049100007861852646, -0.0179038904607296, -0.006025608628988266, -0.06219548359513283, -0.0...
4,015
SelfHVD: Self-Supervised Handheld Video Deblurring
[ "Honglei Xu", "Zhilu Zhang", "Junjie Fan", "Xiaohe Wu", "Wangmeng Zuo" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Xu_SelfHVD_Self-Supervised_Handheld_Video_Deblurring_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Xu_SelfHVD_Self-Supervised_Handheld_Video_Deblurring_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Xu_SelfHVD_Self-Supervised_Handheld_CVPR_2026_supplemental.pdf
2508.08605
cvf
@InProceedings{Xu_2026_CVPR, author = {Xu, Honglei and Zhang, Zhilu and Fan, Junjie and Wu, Xiaohe and Zuo, Wangmeng}, title = {SelfHVD: Self-Supervised Handheld Video Deblurring}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = ...
Shooting video with handheld shooting devices often results in blurry frames due to shaking hands and other instability factors. Although previous video deblurring methods have achieved impressive progress, they still struggle to perform satisfactorily on real-world handheld video due to the blur domain gap between tra...
[ 0.02382122538983822, 0.004043783992528915, -0.01161378063261509, 0.07497522979974747, 0.044631555676460266, 0.0074182068929076195, 0.0026689174119383097, -0.005717014893889427, -0.024564649909734726, -0.04886332154273987, -0.013351491652429104, 0.017536209896206856, -0.02893630787730217, 0...
4,016
RFDM: Residual Flow Diffusion Models for Video Editing
[ "Mohammadreza Salehi", "Mehdi Noroozi", "Luca Morreale", "Ruchika Chavhan", "Malcolm Chadwick", "Alberto Gil Couto Pimentel Ramos", "Abhinav Mehrotra" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Salehi_RFDM_Residual_Flow_Diffusion_Models_for_Video_Editing_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Salehi_RFDM_Residual_Flow_Diffusion_Models_for_Video_Editing_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Salehi_RFDM_Residual_Flow_CVPR_2026_supplemental.pdf
2602.06871
title_judge
@InProceedings{Salehi_2026_CVPR, author = {Salehi, Mohammadreza and Noroozi, Mehdi and Morreale, Luca and Chavhan, Ruchika and Chadwick, Malcolm and Gil Couto Pimentel Ramos, Alberto and Mehrotra, Abhinav}, title = {RFDM: Residual Flow Diffusion Models for Video Editing}, booktitle = {Proceedings of ...
Instructional video editing applies edits to an input video using only text prompts, enabling intuitive natural-language control. Despite the rapid progress, most methods still require fixed-length inputs and substantial compute. Meanwhile, autoregressive video generation enables efficient variable-length synthesis, ye...
[ 0.02240653708577156, -0.004305983893573284, 0.041616663336753845, 0.06113145127892494, 0.0612989217042923, 0.037173978984355927, 0.031148092821240425, 0.005712797865271568, -0.02428733929991722, -0.05723532289266586, -0.006892629899084568, 0.000957159383688122, -0.0245827529579401, -0.0139...
4,017
The Power of Decaying Steps: Enhancing Attack Stability and Transferability for Sign-based Optimizers
[ "Wei Tao", "Yang Dai", "Jincai Huang", "Qing Tao" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Tao_The_Power_of_Decaying_Steps_Enhancing_Attack_Stability_and_Transferability_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Tao_The_Power_of_Decaying_Steps_Enhancing_Attack_Stability_and_Transferability_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Tao_The_Power_of_CVPR_2026_supplemental.pdf
2602.19096
cvf
@InProceedings{Tao_2026_CVPR, author = {Tao, Wei and Dai, Yang and Huang, Jincai and Tao, Qing}, title = {The Power of Decaying Steps: Enhancing Attack Stability and Transferability for Sign-based Optimizers}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognit...
Crafting adversarial examples can be formulated as an optimization problem. While sign-based optimizers such as I-FGSM and MI-FGSM have become the de facto standard for the induced optimization problems, there still exist several unsolved problems in theoretical grounding and practical reliability especially in non-con...
[ -0.01907324232161045, -0.02612285688519478, 0.017516814172267914, 0.045498136430978775, 0.04230552166700363, -0.0032881719525903463, 0.03788163140416145, -0.006787415128201246, -0.0037644279655069113, -0.030620818957686424, 0.009195045568048954, -0.006241480354219675, -0.06318742036819458, ...
4,018
Multi-Scale Gradient-Guided Unrolling Architecture with Adaptive Mamba for Compressive Sensing
[ "Le Yang", "Hongping Gan" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Yang_Multi-Scale_Gradient-Guided_Unrolling_Architecture_with_Adaptive_Mamba_for_Compressive_Sensing_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Yang_Multi-Scale_Gradient-Guided_Unrolling_Architecture_with_Adaptive_Mamba_for_Compressive_Sensing_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Yang_Multi-Scale_Gradient-Guided_Unrolling_CVPR_2026_supplemental.pdf
null
null
@InProceedings{Yang_2026_CVPR, author = {Yang, Le and Gan, Hongping}, title = {Multi-Scale Gradient-Guided Unrolling Architecture with Adaptive Mamba for Compressive Sensing}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = {June...
In the field of Compressive Sensing (CS), deep unrolling networks (DUNs) have demonstrated exceptional performance and interpretability by integrating traditional optimization solvers with deep networks. However, existing DUNs suffer from homogenization in cross-stage feature extraction and insufficient integration of ...
[ -0.023152336478233337, -0.02844400145113468, 0.002193624619394541, 0.02296893112361431, 0.06791006773710251, 0.04882831871509552, 0.023404965177178383, -0.009108876809477806, -0.05333350598812103, -0.07237213104963303, 0.0025170710869133472, -0.014380340464413166, -0.05281408131122589, 0.0...
4,019
OmniVGGT: Omni-Modality Driven Visual Geometry Grounded Transformer
[ "Haosong Peng", "Hao Li", "Yalun Dai", "Yushi Lan", "Yihang Luo", "Tianyu Qi", "Zhengshen Zhang", "Yufeng Zhan", "Junfei Zhang", "Wenchao Xu", "Ziwei Liu" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Peng_OmniVGGT_Omni-Modality_Driven_Visual_Geometry_Grounded_Transformer_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Peng_OmniVGGT_Omni-Modality_Driven_Visual_Geometry_Grounded_Transformer_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Peng_OmniVGGT_Omni-Modality_Driven_CVPR_2026_supplemental.pdf
2511.10560
cvf
@InProceedings{Peng_2026_CVPR, author = {Peng, Haosong and Li, Hao and Dai, Yalun and Lan, Yushi and Luo, Yihang and Qi, Tianyu and Zhang, Zhengshen and Zhan, Yufeng and Zhang, Junfei and Xu, Wenchao and Liu, Ziwei}, title = {OmniVGGT: Omni-Modality Driven Visual Geometry Grounded Transformer}, bookt...
General 3D foundation models have started to lead the trend of unifying diverse vision tasks, yet most assume RGB-only inputs and ignore readily available geometric cues (e.g., camera intrinsics, poses, and depth maps). To address this issue, we introduce OmniVGGT, a novel framework that can effectively benefit from a...
[ 0.026400387287139893, -0.03217710182070732, 0.0377175472676754, 0.032113879919052124, 0.006831924896687269, 0.03778712823987007, 0.02514379657804966, 0.02539171278476715, -0.03492298349738121, -0.050083987414836884, -0.01817389205098152, -0.0047415378503501415, -0.09423379600048065, -0.011...
4,020
Bayesian Decomposition and Semantic Completion for Few-shot Semantic Segmentation
[ "Guangchen Shi", "Yirui Wu", "Wei Zhu", "Tao Wang", "Hao Zhang", "Bo Li", "Tong Lu" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Shi_Bayesian_Decomposition_and_Semantic_Completion_for_Few-shot_Semantic_Segmentation_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Shi_Bayesian_Decomposition_and_Semantic_Completion_for_Few-shot_Semantic_Segmentation_CVPR_2026_paper.pdf
null
null
null
@InProceedings{Shi_2026_CVPR, author = {Shi, Guangchen and Wu, Yirui and Zhu, Wei and Wang, Tao and Zhang, Hao and Li, Bo and Lu, Tong}, title = {Bayesian Decomposition and Semantic Completion for Few-shot Semantic Segmentation}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision ...
Few-shot Semantic Segmentation (FSS) aims to segment objects of novel categories given only a handful of labeled examples. However, existing methods often rely on complex category-specific modeling, resulting in high computational cost and limited generalization under low-data regimes. To address these challenges, we p...
[ 0.014568455517292023, -0.023719487711787224, -0.0034670718014240265, 0.06995942443609238, 0.034598253667354584, 0.03855676203966141, 0.03751732036471367, 0.009574110619723797, -0.05457162857055664, -0.04605313763022423, -0.06315035372972488, 0.0027957865968346596, -0.029609236866235733, -0...
4,021
Natural Human Motion Recovery by Aligning High-Order Temporal Dynamics from Monocular Videos
[ "Dingkun Wei", "Zehong Shen", "Yan Xia", "Georgios Pavlakos", "Yujun Shen", "Xiaowei Zhou" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Wei_Natural_Human_Motion_Recovery_by_Aligning_High-Order_Temporal_Dynamics_from_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Wei_Natural_Human_Motion_Recovery_by_Aligning_High-Order_Temporal_Dynamics_from_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Wei_Natural_Human_Motion_CVPR_2026_supplemental.zip
2605.26879
title_snapshot
@InProceedings{Wei_2026_CVPR, author = {Wei, Dingkun and Shen, Zehong and Xia, Yan and Pavlakos, Georgios and Shen, Yujun and Zhou, Xiaowei}, title = {Natural Human Motion Recovery by Aligning High-Order Temporal Dynamics from Monocular Videos}, booktitle = {Proceedings of the IEEE/CVF Conference on ...
Human motion recovered from monocular videos often appears overly smooth or dynamically inconsistent, even when joint positions are numerically accurate. We observe that this limitation stems from the absence of reliable high-order temporal cues--velocity and acceleration--which are essential for reconstructing motion ...
[ 0.024561293423175812, 0.0032465907279402018, -0.011062469333410263, 0.05061233043670654, 0.016566306352615356, 0.007329403888434172, 0.049005720764398575, 0.023429937660694122, -0.048766568303108215, -0.05391014739871025, -0.005920449271798134, -0.021784333512187004, -0.05504586920142174, ...
4,022
Ego: Embedding-Guided Personalization of Vision-Language Models
[ "Soroush Seifi", "Simon Gardier", "Vaggelis Dorovatas", "Daniel Olmeda Reino", "Rahaf Aljundi" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Seifi_Ego_Embedding-Guided_Personalization_of_Vision-Language_Models_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Seifi_Ego_Embedding-Guided_Personalization_of_Vision-Language_Models_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Seifi_Ego_Embedding-Guided_Personalization_CVPR_2026_supplemental.pdf
2603.09771
cvf
@InProceedings{Seifi_2026_CVPR, author = {Seifi, Soroush and Gardier, Simon and Dorovatas, Vaggelis and Reino, Daniel Olmeda and Aljundi, Rahaf}, title = {Ego: Embedding-Guided Personalization of Vision-Language Models}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Patte...
AI assistants that support humans in daily life are becoming increasingly feasible, driven by the rapid advancements in multimodal language models. A key challenge lies in overcoming the generic nature of these models to deliver personalized experiences. Existing approaches to personalizing large vision language models...
[ 0.03602639213204384, -0.0009901744779199362, 0.018031233921647072, 0.037299320101737976, 0.039808474481105804, 0.05223483592271805, 0.002759996335953474, 0.0163092203438282, -0.010974942706525326, -0.018558908253908157, -0.04565422236919403, 0.021530652418732643, -0.06781227886676788, -0.0...
4,023
Think, Then Verify: A Hypothesis-Verification Multi-Agent Framework for Long Video Understanding
[ "Zheng Wang", "Haoran Chen", "Haoxuan Qin", "Zhipeng Wei", "Tianwen Qian", "Cong Bai" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Wang_Think_Then_Verify_A_Hypothesis-Verification_Multi-Agent_Framework_for_Long_Video_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Wang_Think_Then_Verify_A_Hypothesis-Verification_Multi-Agent_Framework_for_Long_Video_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Wang_Think_Then_Verify_CVPR_2026_supplemental.zip
2603.04977
cvf
@InProceedings{Wang_2026_CVPR, author = {Wang, Zheng and Chen, Haoran and Qin, Haoxuan and Wei, Zhipeng and Qian, Tianwen and Bai, Cong}, title = {Think, Then Verify: A Hypothesis-Verification Multi-Agent Framework for Long Video Understanding}, booktitle = {Proceedings of the IEEE/CVF Conference on ...
Long video understanding is challenging due to dense visual redundancy, long-range temporal dependencies, and the tendency of chain-of-thought and retrieval-based agents to accumulate semantic drift and correlation-driven errors. We argue that long-video reasoning should begin not with reactive retrieval, but with deli...
[ 0.018137892708182335, 0.004357605706900358, -0.01313774287700653, 0.06373758614063263, 0.04405868425965309, 0.005458145402371883, 0.026560606434941292, -0.0037956892047077417, -0.03342945873737335, -0.02986908331513405, -0.014630324207246304, 0.02942214347422123, -0.05305812880396843, -0.0...
4,024
MoVie: Broaden Your Views with Human Motion for Action Detection
[ "Di Yang", "Mahmoud Ali", "Xuanlong Yu", "Xi Shen", "Quan Kong", "Gianpiero Francesca", "François Brémond" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Yang_MoVie_Broaden_Your_Views_with_Human_Motion_for_Action_Detection_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Yang_MoVie_Broaden_Your_Views_with_Human_Motion_for_Action_Detection_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Yang_MoVie_Broaden_Your_CVPR_2026_supplemental.zip
null
null
@InProceedings{Yang_2026_CVPR, author = {Yang, Di and Ali, Mahmoud and Yu, Xuanlong and Shen, Xi and Kong, Quan and Francesca, Gianpiero and Br\'emond, Fran\c{c}ois}, title = {MoVie: Broaden Your Views with Human Motion for Action Detection}, booktitle = {Proceedings of the IEEE/CVF Conference on Com...
Human action detection in videos requires both semantic recognition and accurate modeling of motion. While recent video foundation models have advanced visual semantics, they still struggle to capture complex and compositional actions due to the limited representation ability of motion. Human skeleton sequences, which ...
[ 0.02072622999548912, -0.020834319293498993, -0.012746823020279408, 0.03609121963381767, 0.05177060514688492, 0.017621846869587898, 0.03152427449822426, -0.009317368268966675, -0.056157778948545456, -0.04899165779352188, -0.004418900236487389, -0.013548621907830238, -0.05146961286664009, 0....
4,025
CodeDance: A Dynamic Tool-integrated MLLM for Executable Visual Reasoning
[ "Qi Song", "Honglin Li", "Yingchen Yu", "Haoyi Zhou", "Lin Yang", "Song Bai", "Qi She", "Zilong Huang", "Yunqing Zhao" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Song_CodeDance_A_Dynamic_Tool-integrated_MLLM_for_Executable_Visual_Reasoning_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Song_CodeDance_A_Dynamic_Tool-integrated_MLLM_for_Executable_Visual_Reasoning_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Song_CodeDance_A_Dynamic_CVPR_2026_supplemental.pdf
2512.17312
cvf
@InProceedings{Song_2026_CVPR, author = {Song, Qi and Li, Honglin and Yu, Yingchen and Zhou, Haoyi and Yang, Lin and Bai, Song and She, Qi and Huang, Zilong and Zhao, Yunqing}, title = {CodeDance: A Dynamic Tool-integrated MLLM for Executable Visual Reasoning}, booktitle = {Proceedings of the IEEE/CV...
Recent releases such as o3 highlight human-like "thinking with images" reasoning that combines tool use with stepwise verification, yet most open-source approaches still rely on text-only chains, rigid visual schemas, or single-step pipelines, limiting flexibility, interpretability, and transferability on complex tasks...
[ 0.02700190618634224, -0.03744194656610489, -0.007145341020077467, 0.025128809735178947, 0.048563506454229355, 0.022784925997257233, -0.0045394012704491615, -0.004480823874473572, -0.04046376049518585, -0.03155357018113136, -0.03627288341522217, 0.02964024804532528, -0.06719648838043213, -0...
4,026
SpatialTree: How Spatial Intelligence Branches Out in MLLMs
[ "Yuxi Xiao", "Longfei Li", "Shen Yan", "Xinhang Liu", "Sida Peng", "Yunchao Wei", "Xiaowei Zhou", "Bingyi Kang" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Xiao_SpatialTree_How_Spatial_Intelligence_Branches_Out_in_MLLMs_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Xiao_SpatialTree_How_Spatial_Intelligence_Branches_Out_in_MLLMs_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Xiao_SpatialTree_How_Spatial_CVPR_2026_supplemental.pdf
2512.20617
title_judge
@InProceedings{Xiao_2026_CVPR, author = {Xiao, Yuxi and Li, Longfei and Yan, Shen and Liu, Xinhang and Peng, Sida and Wei, Yunchao and Zhou, Xiaowei and Kang, Bingyi}, title = {SpatialTree: How Spatial Intelligence Branches Out in MLLMs}, booktitle = {Proceedings of the IEEE/CVF Conference on Compute...
Cognitive science suggests that spatial ability develops progressively--from perception to reasoning and interaction. Yet in multimodal LLMs (MLLMs), this hierarchy remains poorly understood, as most studies focus on a narrow set of tasks. We introduce SpatialTree, a cognitive-science-inspired hierarchy that organizes ...
[ 0.0032695713452994823, 0.009472190402448177, 0.032659970223903656, 0.0000729099047021009, 0.04728766158223152, -0.0029236662667244673, 0.040977925062179565, 0.009232885204255581, -0.034786853939294815, -0.0222101379185915, -0.029872892424464226, 0.0017735733417794108, -0.050658825784921646, ...
4,027
HyperST: Hierarchical Hyperbolic Learning for Spatial Transcriptomics Prediction
[ "Chen Zhang", "Yilu An", "Ying Chen", "Hao Li", "Xitong Ling", "Lihao Liu", "Junjun He", "Yuxiang Lin", "Zihui Wang", "Rongshan Yu" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Zhang_HyperST_Hierarchical_Hyperbolic_Learning_for_Spatial_Transcriptomics_Prediction_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Zhang_HyperST_Hierarchical_Hyperbolic_Learning_for_Spatial_Transcriptomics_Prediction_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zhang_HyperST_Hierarchical_Hyperbolic_CVPR_2026_supplemental.pdf
2511.22107
cvf
@InProceedings{Zhang_2026_CVPR, author = {Zhang, Chen and An, Yilu and Chen, Ying and Li, Hao and Ling, Xitong and Liu, Lihao and He, Junjun and Lin, Yuxiang and Wang, Zihui and Yu, Rongshan}, title = {HyperST: Hierarchical Hyperbolic Learning for Spatial Transcriptomics Prediction}, booktitle = {Pro...
Spatial Transcriptomics (ST) merges the benefits of pathology images and gene expression, linking molecular profiles with tissue structure to analyze spot-level function comprehensively. Predicting gene expression from histology images is a cost-effective alternative to expensive ST technologies. However, existing meth...
[ 0.0067428527399897575, 0.008765040896832943, 0.001909398240968585, 0.04373661428689957, 0.05556865409016609, 0.01753188483417034, 0.045803993940353394, 0.005613724701106548, 0.00558899249881506, -0.05962131544947624, -0.0022740361746400595, -0.024574726819992065, -0.04517517611384392, 0.02...
4,028
Towards Robust Sequential Decomposition for Complex Image Editing
[ "Zilai Zeng", "Mingdeng Cao", "Zijie Li", "Xiaochen Lian", "Yichun Shi", "Peihao Zhu", "Chen Sun", "Peng Wang" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Zeng_Towards_Robust_Sequential_Decomposition_for_Complex_Image_Editing_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Zeng_Towards_Robust_Sequential_Decomposition_for_Complex_Image_Editing_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zeng_Towards_Robust_Sequential_CVPR_2026_supplemental.pdf
2605.09233
cvf
@InProceedings{Zeng_2026_CVPR, author = {Zeng, Zilai and Cao, Mingdeng and Li, Zijie and Lian, Xiaochen and Shi, Yichun and Zhu, Peihao and Sun, Chen and Wang, Peng}, title = {Towards Robust Sequential Decomposition for Complex Image Editing}, booktitle = {Proceedings of the IEEE/CVF Conference on Co...
Recent advances in visual generative models have enabled high-fidelity image editing guided by human instructions. However, these models often struggle with complex instructions involving combinatorial editing operations or inter-step dependencies. This difficulty stems from the limitations of two canonical paradigms: ...
[ 0.015250168740749359, 0.0040963576175272465, -0.007077000569552183, 0.0604296512901783, 0.056590329855680466, 0.03291856870055199, 0.021671902388334274, -0.0033688070252537727, -0.018261972814798355, -0.05505616217851639, -0.03319700434803963, -0.0012430682545527816, -0.07382448762655258, ...
4,029
Prototypical Action Reasoning Facilitated by Vision-Language Alignment for Egocentric Action Anticipation
[ "Jiang Shao", "Xinbo Zhao", "Wenyin Tuo", "Xiaochun Zou" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Shao_Prototypical_Action_Reasoning_Facilitated_by_Vision-Language_Alignment_for_Egocentric_Action_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Shao_Prototypical_Action_Reasoning_Facilitated_by_Vision-Language_Alignment_for_Egocentric_Action_CVPR_2026_paper.pdf
null
null
null
@InProceedings{Shao_2026_CVPR, author = {Shao, Jiang and Zhao, Xinbo and Tuo, Wenyin and Zou, Xiaochun}, title = {Prototypical Action Reasoning Facilitated by Vision-Language Alignment for Egocentric Action Anticipation}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Patt...
Egocentric Action Anticipation aims to infer future actions from videos, which is crucial for embodied AI systems. However, its advancement is hindered by the inherent stochasticity of the future, which introduces significant prediction uncertainty. Prevailing methods typically adopt an end-to-end approach to model hol...
[ 0.015438091941177845, -0.00860139075666666, -0.012841868214309216, 0.018470942974090576, 0.009295395575463772, 0.022056177258491516, 0.05675795301795006, -0.0001678605331107974, -0.03693187236785889, -0.018422234803438187, -0.036897990852594376, 0.008203944191336632, -0.05483797937631607, ...
4,030
PECCVAI: Overcoming the Brittleness of AI Image Watermarking Under Visual Paraphrasing Attacks
[ "Shreyas Dixit", "Ashhar Aziz", "Shashwat Bajpai", "Vasu Sharma", "Aman Chadha", "Vinija Jain", "Amitava Das" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Dixit_PECCVAI_Overcoming_the_Brittleness_of_AI_Image_Watermarking_Under_Visual_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Dixit_PECCVAI_Overcoming_the_Brittleness_of_AI_Image_Watermarking_Under_Visual_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Dixit_PECCVAI_Overcoming_the_CVPR_2026_supplemental.pdf
2408.10446
title_judge
@InProceedings{Dixit_2026_CVPR, author = {Dixit, Shreyas and Aziz, Ashhar and Bajpai, Shashwat and Sharma, Vasu and Chadha, Aman and Jain, Vinija and Das, Amitava}, title = {PECCVAI: Overcoming the Brittleness of AI Image Watermarking Under Visual Paraphrasing Attacks}, booktitle = {Proceedings of th...
By 2026, up to 90% of online content may be synthetically generated, raising urgent concerns about the proliferation of AI-driven disinformation. In response, policymakers and technology companies are turning to watermarking as a safeguard: California's Bill AB 321 mandates watermarking of AI-generated media, while fir...
[ 0.03085276670753956, -0.026064323261380196, -0.048476703464984894, 0.0907137468457222, 0.042862311005592346, 0.029294107109308243, 0.02655995823442936, 0.002281567081809044, 0.0012071577366441488, -0.05022376775741577, -0.01993791200220585, -0.0031812849920243025, -0.06420479714870453, 0.0...
4,031
STAC: Plug-and-Play Spatio-Temporal Aware Cache Compression for Streaming 3D Reconstruction
[ "Runze Wang", "Yuxuan Song", "Youcheng Cai", "Ligang Liu" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Wang_STAC_Plug-and-Play_Spatio-Temporal_Aware_Cache_Compression_for_Streaming_3D_Reconstruction_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Wang_STAC_Plug-and-Play_Spatio-Temporal_Aware_Cache_Compression_for_Streaming_3D_Reconstruction_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Wang_STAC_Plug-and-Play_Spatio-Temporal_CVPR_2026_supplemental.pdf
2603.20284
cvf
@InProceedings{Wang_2026_CVPR, author = {Wang, Runze and Song, Yuxuan and Cai, Youcheng and Liu, Ligang}, title = {STAC: Plug-and-Play Spatio-Temporal Aware Cache Compression for Streaming 3D Reconstruction}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogniti...
Online 3D reconstruction from streaming inputs requires both long-term temporal consistency and efficient memory usage. Although causal variants of VGGT address this challenge through a key-value (KV) cache mechanism, the cache grows linearly with the stream length, creating a major memory bottleneck. Under limited mem...
[ 0.00841924361884594, -0.007014505565166473, 0.005107674282044172, 0.023459680378437042, -0.00015687711129430681, 0.0655175969004631, 0.014574555680155754, 0.04451962932944298, -0.028435207903385162, -0.06083757057785988, -0.014031985774636269, -0.032794639468193054, -0.03494216501712799, 0...
4,032
D^3FER: Dual Channel and Dual Branch Network for Robust Facial Expression Recognition under Dual Challenges
[ "Hui Tang", "Yifan He", "Zhong Jin" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Tang_D3FER_Dual_Channel_and_Dual_Branch_Network_for_Robust_Facial_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Tang_D3FER_Dual_Channel_and_Dual_Branch_Network_for_Robust_Facial_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Tang_D3FER_Dual_Channel_CVPR_2026_supplemental.pdf
null
null
@InProceedings{Tang_2026_CVPR, author = {Tang, Hui and He, Yifan and Jin, Zhong}, title = {D{\textasciicircum}3FER: Dual Channel and Dual Branch Network for Robust Facial Expression Recognition under Dual Challenges}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern ...
Facial expression recognition (FER) in the wild is challenged by co-occurring visual perturbations (e.g., occlusions, pose variations) and label noise. Existing methods often address these issues in isolation, failing to handle their compound effects effectively. To this end, we propose D^3FER (Dual channel and Dual br...
[ -0.010860935784876347, -0.01890075020492077, 0.007636144291609526, 0.013809763826429844, 0.021201513707637787, 0.04669709876179695, 0.010443120263516903, -0.012610465288162231, -0.006201352458447218, -0.05595524236559868, -0.003566039027646184, 0.01230165921151638, -0.06964626163244247, 0....
4,033
GA-VLN: Geometry-Aware BEV Representation for Efficient Vision-Language Navigation
[ "Jiahao Yang", "Zihan Wang", "Xiangyang Li", "Xing Zhu", "Yujun Shen", "Yinghao Xu", "Shuqiang Jiang" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Yang_GA-VLN_Geometry-Aware_BEV_Representation_for_Efficient_Vision-Language_Navigation_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Yang_GA-VLN_Geometry-Aware_BEV_Representation_for_Efficient_Vision-Language_Navigation_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Yang_GA-VLN_Geometry-Aware_BEV_CVPR_2026_supplemental.pdf
2605.22036
cvf
@InProceedings{Yang_2026_CVPR, author = {Yang, Jiahao and Wang, Zihan and Li, Xiangyang and Zhu, Xing and Shen, Yujun and Xu, Yinghao and Jiang, Shuqiang}, title = {GA-VLN: Geometry-Aware BEV Representation for Efficient Vision-Language Navigation}, booktitle = {Proceedings of the IEEE/CVF Conference...
Despite significant progress in Vision-Language Navigation (VLN), existing approaches still rely on dense RGB videos that produce excessive patch tokens and lack explicit spatial structure, resulting in substantial computational overhead and limited spatial reasoning. To address these issues, we introduce the Geometry-...
[ 0.00018263496167492121, 0.00798554066568613, 0.022705066949129105, 0.001345034223049879, 0.00752134807407856, 0.039624206721782684, 0.028297919780015945, 0.013317005708813667, -0.026498733088374138, -0.03325363248586655, -0.04254170134663582, 0.010872745886445045, -0.061987414956092834, 0....
4,034
SAME: Sparse and Anchored Model Editing for Heterogeneous Incremental Learning under Limited Data
[ "Zixuan Duan", "Zeyu Zhang", "Fengyuan Lu", "Shaofeng Zhang", "Wenbin Li", "Qi Fan", "Yang Gao" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Duan_SAME_Sparse_and_Anchored_Model_Editing_for_Heterogeneous_Incremental_Learning_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Duan_SAME_Sparse_and_Anchored_Model_Editing_for_Heterogeneous_Incremental_Learning_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Duan_SAME_Sparse_and_CVPR_2026_supplemental.pdf
null
null
@InProceedings{Duan_2026_CVPR, author = {Duan, Zixuan and Zhang, Zeyu and Lu, Fengyuan and Zhang, Shaofeng and Li, Wenbin and Fan, Qi and Gao, Yang}, title = {SAME: Sparse and Anchored Model Editing for Heterogeneous Incremental Learning under Limited Data}, booktitle = {Proceedings of the IEEE/CVF C...
Existing Incremental Learning (IL) methods are primarily evaluated under either a single-domain class-incremental setting, or a multi-domain task-incremental setting with known task identifiers. However, these assumptions often fail to hold in real-world applications. To bridge this gap, we introduce Heterogeneous Incr...
[ -0.008232226595282555, -0.018842823803424835, -0.011896026320755482, 0.034224983304739, 0.034291964024305344, 0.022694841027259827, 0.040757227689027786, -0.00702246418222785, -0.038705479353666306, -0.02074425481259823, -0.00340799312107265, 0.015334437601268291, -0.09100527316331863, -0....
4,035
Context-Nav: Context-Driven Exploration and Viewpoint-Aware 3D Spatial Reasoning for Instance Navigation
[ "Won Shik Jang", "Ue-Hwan Kim" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Jang_Context-Nav_Context-Driven_Exploration_and_Viewpoint-Aware_3D_Spatial_Reasoning_for_Instance_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Jang_Context-Nav_Context-Driven_Exploration_and_Viewpoint-Aware_3D_Spatial_Reasoning_for_Instance_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Jang_Context-Nav_Context-Driven_Exploration_CVPR_2026_supplemental.pdf
2603.09506
cvf
@InProceedings{Jang_2026_CVPR, author = {Jang, Won Shik and Kim, Ue-Hwan}, title = {Context-Nav: Context-Driven Exploration and Viewpoint-Aware 3D Spatial Reasoning for Instance Navigation}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, mo...
Text-goal instance navigation (TGIN) asks an agent to resolve a single, free-form description into actions that reach the correct object instance among same-category distractors. We present Context-Nav, which elevates long, contextual captions from a local matching cue to a global exploration prior and verifies candida...
[ -0.009168347343802452, 0.025443295016884804, 0.005035664886236191, 0.043014589697122574, 0.009065370075404644, 0.014513881877064705, 0.019808677956461906, 0.014836946502327919, -0.03678201511502266, -0.022674694657325745, -0.06334086507558823, 0.0383978933095932, -0.04976736381649971, -0.0...
4,036
QVGGT: Post-Training Quantized Visual Geometry Grounded Transformer
[ "Zhizhen Pan", "Hesong Wang", "Huan Wang" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Pan_QVGGT_Post-Training_Quantized_Visual_Geometry_Grounded_Transformer_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Pan_QVGGT_Post-Training_Quantized_Visual_Geometry_Grounded_Transformer_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Pan_QVGGT_Post-Training_Quantized_CVPR_2026_supplemental.pdf
2509.21302
title_judge
@InProceedings{Pan_2026_CVPR, author = {Pan, Zhizhen and Wang, Hesong and Wang, Huan}, title = {QVGGT: Post-Training Quantized Visual Geometry Grounded Transformer}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = {June}, yea...
Estimating 3D attributes directly from images has advanced rapidly with the Visual Geometry Grounded Transformer (VGGT), which predicts camera parameters, depth maps, and point clouds in a single forward pass. However, its 1.2B-parameter scale severely limits deployment on resource-constrained platforms such as UAVs an...
[ 0.010966514237225056, -0.01792904920876026, 0.0027351845055818558, 0.02686227113008499, 0.016133099794387817, 0.0484088659286499, 0.016137031838297844, -0.006198677234351635, -0.028772586956620216, -0.03929902985692024, -0.03252140060067177, -0.0022342640440911055, -0.09442301094532013, -0...
4,037
OptiMVMap: Offline Vectorized Map Construction via Optimal Multi-vehicle Perspectives
[ "Zedong Dan", "Zijie Wang", "Wei Zhang", "Xiangru Lin", "Weiming Zhang", "Xiao Tan", "Jingdong Wang", "Liang Lin", "Guanbin Li" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Dan_OptiMVMap_Offline_Vectorized_Map_Construction_via_Optimal_Multi-vehicle_Perspectives_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Dan_OptiMVMap_Offline_Vectorized_Map_Construction_via_Optimal_Multi-vehicle_Perspectives_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Dan_OptiMVMap_Offline_Vectorized_CVPR_2026_supplemental.pdf
2604.17135
cvf
@InProceedings{Dan_2026_CVPR, author = {Dan, Zedong and Wang, Zijie and Zhang, Wei and Lin, Xiangru and Zhang, Weiming and Tan, Xiao and Wang, Jingdong and Lin, Liang and Li, Guanbin}, title = {OptiMVMap: Offline Vectorized Map Construction via Optimal Multi-vehicle Perspectives}, booktitle = {Procee...
Offline vectorized maps constitute critical infrastructure for high-precision autonomous driving and mapping services. Existing approaches rely predominantly on single ego-vehicle trajectories, which fundamentally suffer from viewpoint insufficiency: while memory-based methods extend observation time by aggregating ego...
[ -0.003670389298349619, -0.019409721717238426, 0.039363544434309006, 0.0350189134478569, -0.0034914964344352484, 0.05529876798391342, 0.025822661817073822, 0.04296630620956421, -0.04304862767457962, -0.07966526597738266, -0.037260252982378006, -0.01599135994911194, -0.06520330905914307, -0....
4,038
Learning Explicit Continuous Motion Representation for Dynamic Gaussian Splatting from Monocular Videos
[ "Xuankai Zhang", "Junjin Xiao", "Shangwei Huang", "Wei-shi Zheng", "Qing Zhang" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Zhang_Learning_Explicit_Continuous_Motion_Representation_for_Dynamic_Gaussian_Splatting_from_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Zhang_Learning_Explicit_Continuous_Motion_Representation_for_Dynamic_Gaussian_Splatting_from_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zhang_Learning_Explicit_Continuous_CVPR_2026_supplemental.zip
2603.25058
cvf
@InProceedings{Zhang_2026_CVPR, author = {Zhang, Xuankai and Xiao, Junjin and Huang, Shangwei and Zheng, Wei-shi and Zhang, Qing}, title = {Learning Explicit Continuous Motion Representation for Dynamic Gaussian Splatting from Monocular Videos}, booktitle = {Proceedings of the IEEE/CVF Conference on ...
We present an approach for high-quality dynamic Gaussian Splatting from monocular videos. To this end, we in this work go one step further beyond previous methods to explicitly model continuous position and orientation deformation of dynamic Gaussians, using an SE(3) B-spline motion bases with a compact set of control ...
[ 0.00675966264680028, 0.015825605019927025, -0.004080257844179869, 0.05239219218492508, 0.014480557292699814, 0.031887248158454895, 0.028485218062996864, 0.014209468849003315, -0.036493849009275436, -0.05710974335670471, -0.015938596799969673, -0.017706455662846565, -0.06566756963729858, 0....
4,039
Hierarchical Attacks for Multi-Modal Multi-Agent Reasoning
[ "Hao Zhou", "Tiru Wu", "Yan Jiang", "Wanqi Zhou", "Junxing Hu", "Ai Han" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Zhou_Hierarchical_Attacks_for_Multi-Modal_Multi-Agent_Reasoning_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Zhou_Hierarchical_Attacks_for_Multi-Modal_Multi-Agent_Reasoning_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zhou_Hierarchical_Attacks_for_CVPR_2026_supplemental.pdf
2605.13213
cvf
@InProceedings{Zhou_2026_CVPR, author = {Zhou, Hao and Wu, Tiru and Jiang, Yan and Zhou, Wanqi and Hu, Junxing and Han, Ai}, title = {Hierarchical Attacks for Multi-Modal Multi-Agent Reasoning}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, ...
Multi-modal multi-agent systems (MM-MAS) have gained increasing attention for their capacity to enable complex reasoning and coordination across diverse modalities. As these systems continue to expand in scale and functionality, investigating their potential vulnerabilities has become increasingly important.However, ex...
[ -0.019254878163337708, -0.00006522839976241812, -0.0014296001754701138, 0.06397000700235367, 0.02947435900568962, -0.014721041545271873, 0.03779296204447746, 0.006938269827514887, -0.030547454953193665, -0.028533613309264183, -0.009457150474190712, 0.0471627339720726, -0.07677869498729706, ...
4,040
MotionV2V: Editing Motion in a Video
[ "Ryan Burgert", "Charles Herrmann", "Forrester Cole", "Michael S Ryoo", "Neal Wadhwa", "Andrey Voynov", "Nataniel Ruiz" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Burgert_MotionV2V_Editing_Motion_in_a_Video_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Burgert_MotionV2V_Editing_Motion_in_a_Video_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Burgert_MotionV2V_Editing_Motion_CVPR_2026_supplemental.pdf
2511.20640
cvf
@InProceedings{Burgert_2026_CVPR, author = {Burgert, Ryan and Herrmann, Charles and Cole, Forrester and Ryoo, Michael S and Wadhwa, Neal and Voynov, Andrey and Ruiz, Nataniel}, title = {MotionV2V: Editing Motion in a Video}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and P...
While generative video models have achieved remarkable fidelity and consistency, applying these capabilities to video editing remains a complex challenge. Recent research has extensively explored motion controllability as a means to enhance text-to-video generation or image animation; however, we identify precise motio...
[ 0.026771418750286102, -0.0018391276244074106, -0.007533424999564886, 0.04768175631761551, 0.026784515008330345, 0.003958828281611204, 0.05152761936187744, 0.011199945583939552, -0.0337052084505558, -0.06942631304264069, 0.005271593574434519, -0.03269677609205246, -0.04310157522559166, -0.0...
4,041
Human-Centric Multi-Exposure Fusion: Benchmark and Bi-level Cognition Distillation Framework
[ "Jingjie Shang", "Tengyu Ma", "Heng Zhang", "Jinyuan Liu", "Risheng Liu", "Yuan Wang", "Xiaochen Bo" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Shang_Human-Centric_Multi-Exposure_Fusion_Benchmark_and_Bi-level_Cognition_Distillation_Framework_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Shang_Human-Centric_Multi-Exposure_Fusion_Benchmark_and_Bi-level_Cognition_Distillation_Framework_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Shang_Human-Centric_Multi-Exposure_Fusion_CVPR_2026_supplemental.pdf
null
null
@InProceedings{Shang_2026_CVPR, author = {Shang, Jingjie and Ma, Tengyu and Zhang, Heng and Liu, Jinyuan and Liu, Risheng and Wang, Yuan and Bo, Xiaochen}, title = {Human-Centric Multi-Exposure Fusion: Benchmark and Bi-level Cognition Distillation Framework}, booktitle = {Proceedings of the IEEE/CVF ...
Multi-Exposure Fusion (MEF) seeks to generate a single high-quality image from multiple inputs captured at different exposure levels. Despite substantial progress, most existing approaches depend on statistical metrics that poorly reflect human perceptual preferences. Electroencephalography (EEG) provides a direct phys...
[ 0.017197750508785248, 0.01586163230240345, -0.009155876003205776, 0.010806982405483723, 0.025025255978107452, 0.0006338394596241415, 0.02736719325184822, 0.017484599724411964, -0.03958172723650932, -0.05557749792933464, 0.011358246207237244, -0.018669268116354942, -0.07238015532493591, -0....
4,042
DC-Merge: Improving Model Merging with Directional Consistency
[ "Han-Chen Zhang", "Zi-Hao Zhou", "Mao-Lin Luo", "Shimin Di", "Min-Ling Zhang", "Tong Wei" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Zhang_DC-Merge_Improving_Model_Merging_with_Directional_Consistency_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Zhang_DC-Merge_Improving_Model_Merging_with_Directional_Consistency_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zhang_DC-Merge_Improving_Model_CVPR_2026_supplemental.pdf
2603.06242
cvf
@InProceedings{Zhang_2026_CVPR, author = {Zhang, Han-Chen and Zhou, Zi-Hao and Luo, Mao-Lin and Di, Shimin and Zhang, Min-Ling and Wei, Tong}, title = {DC-Merge: Improving Model Merging with Directional Consistency}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern R...
Model merging aims to integrate multiple task-adapted models into a unified model that preserves the knowledge of each task. In this paper, we identify that the key to this knowledge retention lies in maintaining the directional consistency of singular spaces between merged multi-task vector and individual task vectors...
[ -0.0029112021438777447, -0.013548941351473331, -0.023821087554097176, 0.0453997440636158, 0.014324543066322803, 0.02395332232117653, 0.03179585561156273, 0.011194043792784214, -0.046190738677978516, -0.06909667700529099, -0.010378202423453331, -0.0034027532674372196, -0.0963955894112587, -...
4,043
Random Wins All: Rethinking Grouping Strategies for Vision Tokens
[ "Qihang Fan", "Yuang Ai", "Huaibo Huang", "Ran He" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Fan_Random_Wins_All_Rethinking_Grouping_Strategies_for_Vision_Tokens_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Fan_Random_Wins_All_Rethinking_Grouping_Strategies_for_Vision_Tokens_CVPR_2026_paper.pdf
null
2603.00486
cvf
@InProceedings{Fan_2026_CVPR, author = {Fan, Qihang and Ai, Yuang and Huang, Huaibo and He, Ran}, title = {Random Wins All: Rethinking Grouping Strategies for Vision Tokens}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = {June}...
Since Transformers are introduced into vision architectures, their quadratic complexity has always been a significant issue that many research efforts aim to address. A representative approach involves grouping tokens, performing self-attention calculations within each group. To this end, various carefully designed gro...
[ 0.013117402791976929, -0.010406294837594032, 0.01358521543443203, 0.01992192491889, -0.0027872747741639614, 0.041676513850688934, 0.00838971883058548, 0.029386155307292938, -0.04461200162768364, -0.04850931838154793, -0.05457552522420883, -0.008611207827925682, -0.05834176763892174, -0.020...
4,044
R-4B: Incentivizing General-Purpose Auto-Thinking in MLLMs via Bi-Mode Annealing and Reinforce Learning
[ "Qi Yang", "Bolin Ni", "Shiming Xiang", "Houwen Peng" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Yang_R-4B_Incentivizing_General-Purpose_Auto-Thinking_in_MLLMs_via_Bi-Mode_Annealing_and_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Yang_R-4B_Incentivizing_General-Purpose_Auto-Thinking_in_MLLMs_via_Bi-Mode_Annealing_and_CVPR_2026_paper.pdf
null
2508.21113
title_judge
@InProceedings{Yang_2026_CVPR, author = {Yang, Qi and Ni, Bolin and Xiang, Shiming and Peng, Houwen}, title = {R-4B: Incentivizing General-Purpose Auto-Thinking in MLLMs via Bi-Mode Annealing and Reinforce Learning}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern R...
Multimodal Large Language Models (MLLMs) with explicit step-by-step reasoning have achieved strong performance on complex tasks. However, such reasoning is unnecessary for many simple queries and introduces substantial computational overhead. To address this inefficiency, we present R-4B, an auto-thinking MLLM that dyn...
[ -0.015712901949882507, -0.0274765957146883, 0.02477513998746872, 0.026599906384944916, 0.01986335963010788, 0.006182273384183645, 0.023287450894713402, -0.011341191828250885, -0.053844984620809555, 0.00882278848439455, -0.024133533239364624, 0.034399211406707764, -0.06981257349252701, -0.0...
4,045
When Robots Obey the Patch: Universal Transferable Patch Attacks on Vision-Language-Action Models
[ "Hui Lu", "Yi Yu", "Yiming Yang", "Chenyu Yi", "Qixin Zhang", "Bingquan Shen", "Alex C. Kot", "Xudong Jiang" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Lu_When_Robots_Obey_the_Patch_Universal_Transferable_Patch_Attacks_on_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Lu_When_Robots_Obey_the_Patch_Universal_Transferable_Patch_Attacks_on_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Lu_When_Robots_Obey_CVPR_2026_supplemental.pdf
2511.21192
cvf
@InProceedings{Lu_2026_CVPR, author = {Lu, Hui and Yu, Yi and Yang, Yiming and Yi, Chenyu and Zhang, Qixin and Shen, Bingquan and Kot, Alex C. and Jiang, Xudong}, title = {When Robots Obey the Patch: Universal Transferable Patch Attacks on Vision-Language-Action Models}, booktitle = {Proceedings of t...
Vision-Language-Action (VLA) models are vulnerable to adversarial attacks, yet universal and transferable attacks remain underexplored, as most existing patches overfit to a single model and fail in black-box settings. To address this gap, we present a systematic study of universal, transferable adversarial patches aga...
[ 0.010383629240095615, -0.014441332779824734, 0.014007730409502983, 0.024511944502592087, 0.02625131793320179, 0.021124061197042465, 0.03539338707923889, 0.028914738446474075, -0.011236941441893578, -0.01090034656226635, -0.029746906831860542, 0.020304540172219276, -0.0917685404419899, -0.0...
4,046
Life-IQA: Boosting Blind Image Quality Assessment through GCN-enhanced Layer Interaction and MoE-based Feature Decoupling
[ "Long Tang", "Huiyu Duan", "Guoquan Zheng", "Jianbo Zhang", "Jie Hao", "Liang Yuan" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Tang_Life-IQA_Boosting_Blind_Image_Quality_Assessment_through_GCN-enhanced_Layer_Interaction_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Tang_Life-IQA_Boosting_Blind_Image_Quality_Assessment_through_GCN-enhanced_Layer_Interaction_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Tang_Life-IQA_Boosting_Blind_CVPR_2026_supplemental.pdf
2511.19024
title_snapshot
@InProceedings{Tang_2026_CVPR, author = {Tang, Long and Duan, Huiyu and Zheng, Guoquan and Zhang, Jianbo and Hao, Jie and Yuan, Liang}, title = {Life-IQA: Boosting Blind Image Quality Assessment through GCN-enhanced Layer Interaction and MoE-based Feature Decoupling}, booktitle = {Proceedings of the ...
Blind image quality assessment (BIQA) plays a crucial role in evaluating and optimizing visual experience. Most existing BIQA approaches fuse shallow and deep features extracted from backbone networks, while overlooking the unequal contributions to quality prediction. Moreover, while various vision encoder backbones ar...
[ 0.012245705351233482, -0.008210507221519947, 0.026338303461670876, 0.03511941432952881, 0.05753818899393082, 0.02571050450205803, 0.008391628973186016, 0.009909954853355885, -0.008505801670253277, -0.026961714029312134, -0.012197278439998627, 0.025540562346577644, -0.07453689724206924, 0.0...
4,047
Seeing through boxes: Non-Line-of-Sight 3D Reconstruction from Radar Signals
[ "Jiachen Lu", "Hailan Shanbhag", "Haitham Al Hassanieh" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Lu_Seeing_through_boxes_Non-Line-of-Sight_3D_Reconstruction_from_Radar_Signals_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Lu_Seeing_through_boxes_Non-Line-of-Sight_3D_Reconstruction_from_Radar_Signals_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Lu_Seeing_through_boxes_CVPR_2026_supplemental.pdf
2605.29098
title_snapshot
@InProceedings{Lu_2026_CVPR, author = {Lu, Jiachen and Shanbhag, Hailan and Al Hassanieh, Haitham}, title = {Seeing through boxes: Non-Line-of-Sight 3D Reconstruction from Radar Signals}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month...
Reconstructing object geometry from radio frequency (RF) signals is fundamentally challenging due to the lensless imaging nature of RF sensing, which leads to low spatial resolution and high noise. Unlike light signals, RF signals can penetrate occlusions and thus capture information about hidden scenes. Existing Non-L...
[ 0.0021745781414210796, 0.02164296619594097, 0.010088112205266953, -0.006354776676744223, 0.031793203204870224, 0.02706623263657093, -0.00662640668451786, 0.009649447165429592, -0.027182474732398987, -0.07531559467315674, -0.008085194043815136, 0.004507987294346094, -0.03663616627454758, -0...
4,048
Efficient Training for Human Video Generation with Entropy-Guided Prioritized Progressive Learning
[ "Changlin Li", "Jiawei Zhang", "Shuhao Liu", "Sihao Lin", "Zeyi Shi", "Zhihui Li", "Xiaojun Chang" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Li_Efficient_Training_for_Human_Video_Generation_with_Entropy-Guided_Prioritized_Progressive_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Li_Efficient_Training_for_Human_Video_Generation_with_Entropy-Guided_Prioritized_Progressive_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Li_Efficient_Training_for_CVPR_2026_supplemental.pdf
2511.21136
cvf
@InProceedings{Li_2026_CVPR, author = {Li, Changlin and Zhang, Jiawei and Liu, Shuhao and Lin, Sihao and Shi, Zeyi and Li, Zhihui and Chang, Xiaojun}, title = {Efficient Training for Human Video Generation with Entropy-Guided Prioritized Progressive Learning}, booktitle = {Proceedings of the IEEE/CVF...
Human video generation has advanced rapidly with the development of diffusion models, but the high computational cost and substantial memory consumption associated with training these models on high-resolution, multi-frame data pose significant challenges. In this paper, we propose Entropy-Guided Prioritized Progressiv...
[ 0.005530036520212889, -0.02164759673178196, -0.0033348165452480316, 0.04925520718097687, 0.04201440513134003, 0.049570050090551376, 0.01591355726122856, -0.007776706479489803, -0.04207010194659233, -0.06749484688043594, -0.04361046850681305, 0.004036606755107641, -0.04256977140903473, 0.01...
4,049
CoT-Edit: Let CoT Guide Instruction Video Editing
[ "Sen Liang", "Fengbin Guan", "Youliang Zhang", "Xin Li", "Zhibo Chen" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Liang_CoT-Edit_Let_CoT_Guide_Instruction_Video_Editing_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Liang_CoT-Edit_Let_CoT_Guide_Instruction_Video_Editing_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Liang_CoT-Edit_Let_CoT_CVPR_2026_supplemental.pdf
null
null
@InProceedings{Liang_2026_CVPR, author = {Liang, Sen and Guan, Fengbin and Zhang, Youliang and Li, Xin and Chen, Zhibo}, title = {CoT-Edit: Let CoT Guide Instruction Video Editing}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month =...
Text-driven instruction-based video editing in complex scenes remains challenging: purely textual prompts often fail to capture precise spatial relationships and physical constraints, resulting in target ambiguity and physically implausible outcomes. To address this, we propose a plan-guide-edit framework that explicit...
[ 0.023788943886756897, 0.013192676939070225, -0.016688868403434753, 0.05651751905679703, 0.04575036093592644, 0.008657524362206459, 0.012892180122435093, 0.014512031339108944, -0.020352443680167198, -0.046597495675086975, -0.06180913746356964, 0.022015253081917763, -0.06868553906679153, -0....
4,050
Sparse Spectral LoRA: Routed Experts for Medical VLMs
[ "Omid Nejatimanzari", "Hojat Asgariandehkordi", "Taha Koleilat", "Yiming Xiao", "Hassan Rivaz" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Nejatimanzari_Sparse_Spectral_LoRA_Routed_Experts_for_Medical_VLMs_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Nejatimanzari_Sparse_Spectral_LoRA_Routed_Experts_for_Medical_VLMs_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Nejatimanzari_Sparse_Spectral_LoRA_CVPR_2026_supplemental.pdf
2604.01310
title_snapshot
@InProceedings{Nejatimanzari_2026_CVPR, author = {Nejatimanzari, Omid and Asgariandehkordi, Hojat and Koleilat, Taha and Xiao, Yiming and Rivaz, Hassan}, title = {Sparse Spectral LoRA: Routed Experts for Medical VLMs}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern...
Large vision-language models (VLMs) excel on general benchmarks but often lack robustness in medical imaging, where heterogeneous supervision induces cross-dataset interference and sensitivity to data regime (i.e., how the supervisory signals are mixed). In realistic clinical workflows, data and tasks arrive sequential...
[ 0.01743515022099018, -0.032951291650533676, 0.004319943953305483, 0.0038912231102585793, 0.021771153435111046, 0.004591470118612051, 0.05007468909025192, -0.02471281588077545, -0.03642614558339119, -0.06266748905181885, -0.00017584938905201852, 0.04028989002108574, -0.06461485475301743, 0....
4,051
MFEN: Multi-Frequency Expert Network for Visible-Infrared Person Re-ID
[ "Xulin Li", "Yan Lu", "Bin Liu", "Qinhong Yang", "Qi Chu", "Tao Gong", "Nenghai Yu" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Li_MFEN_Multi-Frequency_Expert_Network_for_Visible-Infrared_Person_Re-ID_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Li_MFEN_Multi-Frequency_Expert_Network_for_Visible-Infrared_Person_Re-ID_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Li_MFEN_Multi-Frequency_Expert_CVPR_2026_supplemental.pdf
null
null
@InProceedings{Li_2026_CVPR, author = {Li, Xulin and Lu, Yan and Liu, Bin and Yang, Qinhong and Chu, Qi and Gong, Tao and Yu, Nenghai}, title = {MFEN: Multi-Frequency Expert Network for Visible-Infrared Person Re-ID}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern ...
Visible-infrared person re-identification (VI-ReID) is challenging due to the large modality discrepancy between visible and infrared images. We contend that this discrepancy is largely related to differing lighting conditions, including differences in light wavelength and light source type. Recently, frequency-based V...
[ 0.03288950026035309, 0.00029356175218708813, 0.03512417525053024, 0.017414171248674393, 0.0468982495367527, 0.018785839900374413, 0.03756674379110336, -0.022467676550149918, -0.035574641078710556, -0.07735034078359604, -0.04195772111415863, 0.04153762757778168, -0.09160807728767395, 0.0000...
4,052
UniRain: Unified Image Deraining with RAG-based Dataset Distillation and Multi-objective Reweighted Optimization
[ "Qianfeng Yang", "Qiyuan Guan", "Xiang Chen", "Jiyu Jin", "Guiyue Jin", "Jiangxin Dong" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Yang_UniRain_Unified_Image_Deraining_with_RAG-based_Dataset_Distillation_and_Multi-objective_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Yang_UniRain_Unified_Image_Deraining_with_RAG-based_Dataset_Distillation_and_Multi-objective_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Yang_UniRain_Unified_Image_CVPR_2026_supplemental.pdf
2603.03967
cvf
@InProceedings{Yang_2026_CVPR, author = {Yang, Qianfeng and Guan, Qiyuan and Chen, Xiang and Jin, Jiyu and Jin, Guiyue and Dong, Jiangxin}, title = {UniRain: Unified Image Deraining with RAG-based Dataset Distillation and Multi-objective Reweighted Optimization}, booktitle = {Proceedings of the IEEE/...
Despite significant progress has been made in image deraining, we note that most existing methods are often developed for only specific types of rain degradation and fail to generalize across diverse real-world rainy scenes. How to effectively model different rain degradations within a universal framework is important ...
[ 0.022706864401698112, -0.04050587862730026, 0.011312603950500488, 0.05837476626038551, 0.028972679749131203, -0.014167604967951775, 0.023185350000858307, -0.01921333372592926, -0.028471825644373894, -0.03781851753592491, -0.03209317475557327, -0.0037440962623804808, -0.06351082772016525, 0...
4,053
From Softmax to Dirichlet: Evidential Learning for Semi-supervised Semantic Segmentation
[ "Huayu Mai", "Rui Sun", "Yujia Chen", "Wangkai Li", "Bingzhou Wang", "Aibing Li", "Zhangyu He", "Yuan Wang" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Mai_From_Softmax_to_Dirichlet_Evidential_Learning_for_Semi-supervised_Semantic_Segmentation_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Mai_From_Softmax_to_Dirichlet_Evidential_Learning_for_Semi-supervised_Semantic_Segmentation_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Mai_From_Softmax_to_CVPR_2026_supplemental.pdf
null
null
@InProceedings{Mai_2026_CVPR, author = {Mai, Huayu and Sun, Rui and Chen, Yujia and Li, Wangkai and Wang, Bingzhou and Li, Aibing and He, Zhangyu and Wang, Yuan}, title = {From Softmax to Dirichlet: Evidential Learning for Semi-supervised Semantic Segmentation}, booktitle = {Proceedings of the IEEE/C...
The critical challenge of semi-supervised semantic segmentation lies in how to fully exploit a large volume of unlabeled data to improve the model's generalization performance for robust segmentation. However, existing softmax scores-based filtering methods tend to be affected by the overconfidence issue in neural netw...
[ -0.011847051791846752, -0.04538441449403763, 0.025847723707556725, 0.048682600259780884, 0.00550106493756175, 0.003862364450469613, 0.014745821245014668, -0.011551282368600368, -0.024984650313854218, -0.05708303675055504, -0.049735672771930695, 0.012795260176062584, -0.057035841047763824, ...
4,054
FlowDirector: Training-Free Flow Steering for Precise Text-to-Video Editing
[ "Guangzhao Li", "Yanming Yang", "Chenxi Song", "Xiaohong Liu", "Chi Zhang" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Li_FlowDirector_Training-Free_Flow_Steering_for_Precise_Text-to-Video_Editing_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Li_FlowDirector_Training-Free_Flow_Steering_for_Precise_Text-to-Video_Editing_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Li_FlowDirector_Training-Free_Flow_CVPR_2026_supplemental.zip
2506.05046
cvf
@InProceedings{Li_2026_CVPR, author = {Li, Guangzhao and Yang, Yanming and Song, Chenxi and Liu, Xiaohong and Zhang, Chi}, title = {FlowDirector: Training-Free Flow Steering for Precise Text-to-Video Editing}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognit...
Text-driven video editing aims to modify video content based on natural language instructions. While recent training-free methods have leveraged pretrained diffusion models, they often rely on an inversion-editing paradigm. This paradigm maps the video to a latent space before editing. However, the inversion process is...
[ 0.00933645386248827, -0.013020230457186699, -0.00720433983951807, 0.04736996442079544, 0.028815824538469315, 0.021652767434716225, 0.031193891540169716, 0.027073219418525696, -0.017081765457987785, -0.05870346352458, -0.018877672031521797, -0.003972186706960201, -0.05158475041389465, 0.002...
4,055
ImageRAGTurbo: Towards One-step Text-to-Image Generation with Retrieval-Augmented Diffusion Models
[ "Peijie Qiu", "Hariharan Ramshankar", "Arnau Ramisa", "Amit Kumar K C", "René Vidal", "Vamsi Salaka", "Rahul Bhagat" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Qiu_ImageRAGTurbo_Towards_One-step_Text-to-Image_Generation_with_Retrieval-Augmented_Diffusion_Models_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Qiu_ImageRAGTurbo_Towards_One-step_Text-to-Image_Generation_with_Retrieval-Augmented_Diffusion_Models_CVPR_2026_paper.pdf
null
2602.12640
cvf
@InProceedings{Qiu_2026_CVPR, author = {Qiu, Peijie and Ramshankar, Hariharan and Ramisa, Arnau and C, Amit Kumar K and Vidal, Ren\'e and Salaka, Vamsi and Bhagat, Rahul}, title = {ImageRAGTurbo: Towards One-step Text-to-Image Generation with Retrieval-Augmented Diffusion Models}, booktitle = {Procee...
Diffusion models have emerged as the leading approach for text-to-image generation. However, their iterative sampling process, which gradually morphs random noise into coherent images, introduces significant latency that limits their applicability. While recent few-step diffusion models reduce the number of sampling st...
[ 0.0036209900863468647, -0.056680697947740555, -0.0005302992649376392, 0.07473620772361755, 0.04103383049368858, -0.019606605172157288, 0.02236255258321762, 0.01669023372232914, -0.024291938170790672, -0.041779886931180954, -0.03137744218111038, -0.009160284884274006, -0.05521334707736969, ...
4,056
Principled Steering via Null-space Projection for Jailbreak Defense in Vision-Language Models
[ "Xingyu Zhu", "Beier Zhu", "Shuo Wang", "Junfeng Fang", "Kesen Zhao", "Hanwang Zhang", "Xiangnan He" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Zhu_Principled_Steering_via_Null-space_Projection_for_Jailbreak_Defense_in_Vision-Language_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Zhu_Principled_Steering_via_Null-space_Projection_for_Jailbreak_Defense_in_Vision-Language_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zhu_Principled_Steering_via_CVPR_2026_supplemental.pdf
2603.22094
cvf
@InProceedings{Zhu_2026_CVPR, author = {Zhu, Xingyu and Zhu, Beier and Wang, Shuo and Fang, Junfeng and Zhao, Kesen and Zhang, Hanwang and He, Xiangnan}, title = {Principled Steering via Null-space Projection for Jailbreak Defense in Vision-Language Models}, booktitle = {Proceedings of the IEEE/CVF C...
As vision-language models (VLMs) are increasingly deployed in open-world scenarios, they can be easily induced by visual jailbreak attacks to generate harmful content, posing serious risks to model safety and trustworthy usage.Recent activation steering methods inject directional vectors into model activations during i...
[ -0.029270995408296585, -0.028107238933444023, -0.007964669726788998, 0.03654865920543671, 0.019562313333153725, 0.02777094393968582, 0.06176981329917908, -0.02639208361506462, -0.03823050856590271, -0.010159127414226532, -0.04525067284703255, 0.011940493248403072, -0.06974852830171585, -0....
4,057
DiT-IC: Aligned Diffusion Transformer for Efficient Image Compression
[ "Junqi Shi", "Ming Lu", "Xingchen Li", "Anle Ke", "Ruiqi Zhang", "Zhan Ma" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Shi_DiT-IC_Aligned_Diffusion_Transformer_for_Efficient_Image_Compression_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Shi_DiT-IC_Aligned_Diffusion_Transformer_for_Efficient_Image_Compression_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Shi_DiT-IC_Aligned_Diffusion_CVPR_2026_supplemental.pdf
2603.13162
cvf
@InProceedings{Shi_2026_CVPR, author = {Shi, Junqi and Lu, Ming and Li, Xingchen and Ke, Anle and Zhang, Ruiqi and Ma, Zhan}, title = {DiT-IC: Aligned Diffusion Transformer for Efficient Image Compression}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition...
Diffusion-based image compression has recently shown outstanding perceptual fidelity, yet its practicality is hindered by prohibitive sampling overhead and high memory usage.Most existing diffusion codecs employ UNet architectures, where hierarchical downsampling forces diffusion to operate in shallow latent spaces (ty...
[ 0.0014501848490908742, -0.015391398221254349, -0.03406954184174538, 0.056504473090171814, 0.06079980731010437, 0.042500972747802734, 0.0005737180472351611, 0.007982472889125347, -0.0049042245373129845, -0.05823618918657303, 0.019824417307972908, -0.0327310785651207, -0.026734692975878716, ...
4,058
PointTPA: Dynamic Network Parameter Adaptation for 3D Scene Understanding
[ "Siyuan Liu", "Chaoqun Zheng", "Xin Zhou", "Tianrui Feng", "Dingkang Liang", "Xiang Bai" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Liu_PointTPA_Dynamic_Network_Parameter_Adaptation_for_3D_Scene_Understanding_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Liu_PointTPA_Dynamic_Network_Parameter_Adaptation_for_3D_Scene_Understanding_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Liu_PointTPA_Dynamic_Network_CVPR_2026_supplemental.pdf
2604.04933
cvf
@InProceedings{Liu_2026_CVPR, author = {Liu, Siyuan and Zheng, Chaoqun and Zhou, Xin and Feng, Tianrui and Liang, Dingkang and Bai, Xiang}, title = {PointTPA: Dynamic Network Parameter Adaptation for 3D Scene Understanding}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and P...
Scene-level point cloud understanding remains challenging due to diverse geometries, imbalanced category distributions, and highly varied spatial layouts. Existing methods improve object-level performance but rely on static network parameters during inference, limiting their adaptability to dynamic scene data. We propo...
[ 0.011216011829674244, -0.007843182422220707, 0.02627135068178177, 0.02582768350839615, -0.003553947899490595, 0.048381004482507706, 0.012678474187850952, -0.01119564101099968, -0.02748037874698639, -0.04719715937972069, -0.02185143157839775, -0.03443940356373787, -0.07197041809558868, 0.00...
4,059
Can Natural Image Autoencoders Compactly Tokenize fMRI Volumes for Long-Range Dynamics Modeling?
[ "Peter Yongho Kim", "Juhyeon Park", "Jungwoo Park", "Jubin Choi", "Jungwoo Seo", "Jiook Cha", "Taesup Moon" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Kim_Can_Natural_Image_Autoencoders_Compactly_Tokenize_fMRI_Volumes_for_Long-Range_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Kim_Can_Natural_Image_Autoencoders_Compactly_Tokenize_fMRI_Volumes_for_Long-Range_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Kim_Can_Natural_Image_CVPR_2026_supplemental.pdf
2604.03619
cvf
@InProceedings{Kim_2026_CVPR, author = {Kim, Peter Yongho and Park, Juhyeon and Park, Jungwoo and Choi, Jubin and Seo, Jungwoo and Cha, Jiook and Moon, Taesup}, title = {Can Natural Image Autoencoders Compactly Tokenize fMRI Volumes for Long-Range Dynamics Modeling?}, booktitle = {Proceedings of the ...
Modeling long-range spatiotemporal dynamics in functional Magnetic Resonance Imaging (fMRI) remains a key challenge due to the high dimensionality of the four-dimensional signals. Prior voxel-based models, although demonstrating excellent performance and interpretation capabilities, are constrained by prohibitive memor...
[ -0.0029517749790102243, -0.030965100973844528, -0.0008090024930424988, 0.014878597110509872, 0.01367881242185831, 0.054692842066287994, 0.038458649069070816, 0.009988012723624706, -0.032668378204107285, -0.03779502585530281, 0.001931702601723373, -0.03604040667414665, -0.047449905425310135, ...
4,060
Dexterous World Models
[ "Byungjun Kim", "Taeksoo Kim", "Junyoung Lee", "Hanbyul Joo" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Kim_Dexterous_World_Models_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Kim_Dexterous_World_Models_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Kim_Dexterous_World_Models_CVPR_2026_supplemental.pdf
2512.17907
cvf
@InProceedings{Kim_2026_CVPR, author = {Kim, Byungjun and Kim, Taeksoo and Lee, Junyoung and Joo, Hanbyul}, title = {Dexterous World Models}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = {June}, year = {2026}, pag...
Recent progress in 3D reconstruction has made it easy to create realistic digital twins from everyday environments. However, current digital twins remain largely static--limited to navigation and view synthesis without embodied interactivity. To bridge this gap, we introduce Dexterous World Model (DWM), an scene-action...
[ 0.004601854365319014, 0.01792645826935768, 0.004986172076314688, 0.03362959623336792, 0.049497026950120926, 0.03177788481116295, 0.02393498085439205, 0.02233370766043663, -0.03092707321047783, -0.05614208057522774, -0.017936551943421364, -0.018969427794218063, -0.05412079766392708, -0.0182...
4,061
You Only Erase Once: Erasing Anything without Bringing Unexpected Content
[ "Yixing Zhu", "Qing Zhang", "Wenju Xu", "Wei-Shi Zheng" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Zhu_You_Only_Erase_Once_Erasing_Anything_without_Bringing_Unexpected_Content_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Zhu_You_Only_Erase_Once_Erasing_Anything_without_Bringing_Unexpected_Content_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Zhu_You_Only_Erase_CVPR_2026_supplemental.pdf
2603.27599
cvf
@InProceedings{Zhu_2026_CVPR, author = {Zhu, Yixing and Zhang, Qing and Xu, Wenju and Zheng, Wei-Shi}, title = {You Only Erase Once: Erasing Anything without Bringing Unexpected Content}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month...
We present YOEO, an approach for object erasure. Unlike recent diffusion-based methods which struggle to erase target objects without generating unexpected content within the masked regions due to lack of sufficient paired training data and explicit constraint on content generation, our method allows to produce high-qu...
[ -0.010029131546616554, 0.016403205692768097, -0.009473230689764023, 0.07227826118469238, 0.037422023713588715, 0.005620649550110102, 0.02704647183418274, 0.007739668246358633, -0.05754277855157852, -0.03250562399625778, -0.010130541399121284, -0.014669213443994522, -0.047560915350914, -0.0...
4,062
TokenGS: Decoupling 3D Gaussian Prediction from Pixels with Learnable Tokens
[ "Jiawei Ren", "Michal Jan Tyszkiewicz", "Jiahui Huang", "Zan Gojcic" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Ren_TokenGS_Decoupling_3D_Gaussian_Prediction_from_Pixels_with_Learnable_Tokens_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Ren_TokenGS_Decoupling_3D_Gaussian_Prediction_from_Pixels_with_Learnable_Tokens_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Ren_TokenGS_Decoupling_3D_CVPR_2026_supplemental.zip
2604.15239
cvf
@InProceedings{Ren_2026_CVPR, author = {Ren, Jiawei and Tyszkiewicz, Michal Jan and Huang, Jiahui and Gojcic, Zan}, title = {TokenGS: Decoupling 3D Gaussian Prediction from Pixels with Learnable Tokens}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (C...
In this work, we revisit several key design choices of modern Transformer-based approaches for feed-forward 3D Gaussian Splatting (3DGS) prediction. We argue that the common practice of regressing Gaussian means as depths along camera rays is suboptimal, and instead propose to directly regress 3D mean coordinates using...
[ 0.027519110590219498, -0.008994650095701218, 0.022677039727568626, 0.04300225153565407, 0.013582310639321804, 0.057425327599048615, 0.008440776728093624, 0.011158201843500137, -0.043499141931533813, -0.04603321850299835, -0.010612370446324348, -0.025245944038033485, -0.06310854852199554, -...
4,063
EditMGT: Unleashing Potentials of Masked Generative Transformers in Image Editing
[ "Wei Chow", "Linfeng Li", "Lingdong Kong", "Zefeng Li", "Qi Xu", "Hang Song", "Tian Ye", "Xian Wang", "Jinbin Bai", "Shilin Xu", "Xiangtai Li", "Junting Pan", "Shaoteng Liu", "Ran Zhou", "Tianshu Yang", "Songhua Liu" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Chow_EditMGT_Unleashing_Potentials_of_Masked_Generative_Transformers_in_Image_Editing_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Chow_EditMGT_Unleashing_Potentials_of_Masked_Generative_Transformers_in_Image_Editing_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Chow_EditMGT_Unleashing_Potentials_CVPR_2026_supplemental.pdf
2512.11715
cvf
@InProceedings{Chow_2026_CVPR, author = {Chow, Wei and Li, Linfeng and Kong, Lingdong and Li, Zefeng and Xu, Qi and Song, Hang and Ye, Tian and Wang, Xian and Bai, Jinbin and Xu, Shilin and Li, Xiangtai and Pan, Junting and Liu, Shaoteng and Zhou, Ran and Yang, Tianshu and Liu, Songhua}, title = {EditMGT...
Recent advances in diffusion models (DMs) have achieved exceptional visual quality in image editing tasks. However, the global denoising dynamics of DMs inherently conflate local editing targets with the full-image context, leading to unintended modifications in non-target regions. In this paper, we shift our attention...
[ 0.0032723441254347563, -0.0017539148684591055, -0.006036778911948204, 0.07810040563344955, 0.04578525573015213, 0.035085562616586685, 0.03856050595641136, 0.012255722656846046, -0.02685515768826008, -0.04795337840914726, -0.039469122886657715, -0.0032957387156784534, -0.0451614148914814, -...
4,064
World in a Frame: Understanding Culture Mixing as a New Challenge for Vision-Language Models
[ "Eunsu Kim", "Junyeong Park", "Na Min An", "Junseong Kim", "Hitesh Laxmichand Patel", "Jiho Jin", "Julia Kruk", "Amit Agarwal", "Srikant Panda", "Fenal Ashokbhai Ilasariya", "Hyunjung Shim", "Alice Oh" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Kim_World_in_a_Frame_Understanding_Culture_Mixing_as_a_New_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Kim_World_in_a_Frame_Understanding_Culture_Mixing_as_a_New_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Kim_World_in_a_CVPR_2026_supplemental.pdf
2511.22787
cvf
@InProceedings{Kim_2026_CVPR, author = {Kim, Eunsu and Park, Junyeong and An, Na Min and Kim, Junseong and Patel, Hitesh Laxmichand and Jin, Jiho and Kruk, Julia and Agarwal, Amit and Panda, Srikant and Ilasariya, Fenal Ashokbhai and Shim, Hyunjung and Oh, Alice}, title = {World in a Frame: Understanding...
In a globalized world, cultural elements from diverse origins frequently appear together within a single visual scene. We refer to these as culture mixing scenarios, yet how Large Vision-Language Models (LVLMs) perceive them remains underexplored. We investigate culture mixing as a critical challenge for LVLMs and exam...
[ 0.044321708381175995, 0.020955445244908333, -0.018490275368094444, 0.05516691133379936, 0.02517188899219036, 0.016781900078058243, 0.021170638501644135, 0.04773861542344093, -0.03376119211316109, -0.0275726281106472, -0.02673034369945526, 0.015393638052046299, -0.061303023248910904, -0.002...
4,065
NuWa: Deriving Lightweight Class-Specific Vision Transformers for Edge Devices
[ "Ziteng Wei", "Qiang He", "Bing Li", "Feifei Chen", "Hai Jin", "Yun Yang" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Wei_NuWa_Deriving_Lightweight_Class-Specific_Vision_Transformers_for_Edge_Devices_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Wei_NuWa_Deriving_Lightweight_Class-Specific_Vision_Transformers_for_Edge_Devices_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Wei_NuWa_Deriving_Lightweight_CVPR_2026_supplemental.zip
2504.03118
title_judge
@InProceedings{Wei_2026_CVPR, author = {Wei, Ziteng and He, Qiang and Li, Bing and Chen, Feifei and Jin, Hai and Yang, Yun}, title = {NuWa: Deriving Lightweight Class-Specific Vision Transformers for Edge Devices}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Rec...
Vision Transformers (ViTs) often need to be compressed for deployment on resource-constrained edge devices like drones and smart vehicles. However, existing model compression methods ignore that many edge devices only require the knowledge of specific classes for their applications. As a result, the derived all-class V...
[ -0.004517252556979656, -0.04596366360783577, -0.018150901421904564, 0.0315074548125267, 0.015127249993383884, 0.03851384297013283, 0.019651448354125023, 0.00011341857316438109, 0.004806023556739092, -0.045424312353134155, -0.017046252265572548, 0.026633992791175842, -0.06234043464064598, 0...
4,066
SCAPO: Self-Supervised Category-Level Articulated Pose Estimation from a Single 3D Observation
[ "Can Zhang", "Gim Hee Lee" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Zhang_SCAPO_Self-Supervised_Category-Level_Articulated_Pose_Estimation_from_a_Single_3D_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Zhang_SCAPO_Self-Supervised_Category-Level_Articulated_Pose_Estimation_from_a_Single_3D_CVPR_2026_paper.pdf
null
null
null
@InProceedings{Zhang_2026_CVPR, author = {Zhang, Can and Lee, Gim Hee}, title = {SCAPO: Self-Supervised Category-Level Articulated Pose Estimation from a Single 3D Observation}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = {Ju...
Existing methods for category-level object articulation from a single 3D observation often rely on dense supervision, multi-frame inputs, or CAD templates, and still struggle to disentangle geometry from articulation or to recover explicit joint parameters. We propose SCAPO , a self-supervised framework that estimates ...
[ 0.03120555728673935, -0.010872792452573776, -0.037035416811704636, 0.011600364930927753, 0.02411358430981636, 0.06606493890285492, 0.011945060454308987, -0.006222285330295563, -0.05483691766858101, -0.047747328877449036, -0.006853681523352861, -0.024853145703673363, -0.09199023991823196, -...
4,067
CGHair: Compact Gaussian Hair Reconstruction with Card Clustering
[ "Haimin Luo", "Srinjay Sarkar", "Albert Mosella-Montoro", "Francisco Vicente Carrasco", "Fernando De la Torre" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Luo_CGHair_Compact_Gaussian_Hair_Reconstruction_with_Card_Clustering_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Luo_CGHair_Compact_Gaussian_Hair_Reconstruction_with_Card_Clustering_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Luo_CGHair_Compact_Gaussian_CVPR_2026_supplemental.zip
2604.03716
title_snapshot
@InProceedings{Luo_2026_CVPR, author = {Luo, Haimin and Sarkar, Srinjay and Mosella-Montoro, Albert and Carrasco, Francisco Vicente and De la Torre, Fernando}, title = {CGHair: Compact Gaussian Hair Reconstruction with Card Clustering}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer ...
We present a compact pipeline for high-fidelity hair reconstruction from multi-view images. While recent 3D Gaussian Splatting (3DGS) methods achieve realistic results, they often require millions of primitives, leading to high storage and rendering costs. Observing that hair exhibits structural and visual similarities...
[ 0.027100132778286934, -0.023966597393155098, 0.007849503308534622, 0.046885691583156586, 0.022712567821145058, 0.027243895456194878, 0.014363477006554604, 0.0034807792399078608, -0.016947230324149132, -0.06944742053747177, -0.003178682178258896, -0.024320203810930252, -0.06303689628839493, ...
4,068
Frequency-domain Manipulation for Face Obfuscation
[ "Jintae Kim", "Keunsoo Ko", "Chang-Su Kim" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Kim_Frequency-domain_Manipulation_for_Face_Obfuscation_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Kim_Frequency-domain_Manipulation_for_Face_Obfuscation_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Kim_Frequency-domain_Manipulation_for_CVPR_2026_supplemental.zip
null
null
@InProceedings{Kim_2026_CVPR, author = {Kim, Jintae and Ko, Keunsoo and Kim, Chang-Su}, title = {Frequency-domain Manipulation for Face Obfuscation}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = {June}, year = {2026},...
Facial image datasets have become essential resources for various face analysis tasks, but their use raises significant privacy concerns. To address this issue, face obfuscation has emerged as a practical approach to hide identity from humans while retaining cues decipherable by machines. However, existing methods ofte...
[ -0.013159876689314842, 0.0057409098371863365, -0.005071117542684078, 0.018470553681254387, 0.04532996937632561, 0.0066309962421655655, 0.04366351291537285, -0.023236531764268875, -0.011806335300207138, -0.03324010595679283, -0.03205297142267227, 0.041902460157871246, -0.0543980672955513, -...
4,069
E3AD: An Emotion-Aware Vision-Language-Action Model for Human-Centric End-to-End Autonomous Driving
[ "Yihong Tang", "Haicheng Liao", "Tong Nie", "Junlin He", "Ao Qu", "Kehua Chen", "Wei Ma", "Zhenning Li", "Lijun Sun", "Chengzhong Xu" ]
https://openaccess.thecvf.com/content/CVPR2026/html/Tang_E3AD_An_Emotion-Aware_Vision-Language-Action_Model_for_Human-Centric_End-to-End_Autonomous_Driving_CVPR_2026_paper.html
https://openaccess.thecvf.com/content/CVPR2026/papers/Tang_E3AD_An_Emotion-Aware_Vision-Language-Action_Model_for_Human-Centric_End-to-End_Autonomous_Driving_CVPR_2026_paper.pdf
https://openaccess.thecvf.com/content/CVPR2026/supplemental/Tang_E3AD_An_Emotion-Aware_CVPR_2026_supplemental.pdf
2512.04733
cvf
@InProceedings{Tang_2026_CVPR, author = {Tang, Yihong and Liao, Haicheng and Nie, Tong and He, Junlin and Qu, Ao and Chen, Kehua and Ma, Wei and Li, Zhenning and Sun, Lijun and Xu, Chengzhong}, title = {E3AD: An Emotion-Aware Vision-Language-Action Model for Human-Centric End-to-End Autonomous Driving}, ...
End-to-end autonomous driving (AD) systems increasingly adopt vision-language-action (VLA) models, yet they ignore the passenger's emotional state, which is central to comfort and AD acceptance. We introduce Open-Domain End-to-End (OD-E2E) AD, where an autonomous vehicle must interpret free-form natural-language comman...
[ -0.008076466619968414, 0.012983419932425022, 0.02458980493247509, 0.04288417100906372, -0.0007761840824969113, 0.032987870275974274, 0.017824172973632812, 0.027670789510011673, 0.0030441421549767256, -0.042001865804195404, -0.024447964504361153, 0.027938516810536385, -0.05587029829621315, ...