Henry Fu
henryfu-at-bd
AI & ML interests
None yet
Recent Activity
updated a collection about 2 months ago
Diffusion Models updated a collection 3 months ago
AR Video Generation updated a collection 4 months ago
AR Video GenerationOrganizations
AR Video Generation
-
Deep Forcing: Training-Free Long Video Generation with Deep Sink and Participative Compression
Paper • 2512.05081 • Published • 33 -
Reward Forcing: Efficient Streaming Video Generation with Rewarded Distribution Matching Distillation
Paper • 2512.04678 • Published • 42 -
What about gravity in video generation? Post-Training Newton's Laws with Verifiable Rewards
Paper • 2512.00425 • Published • 53 -
Flash-DMD: Towards High-Fidelity Few-Step Image Generation with Efficient Distillation and Joint Reinforcement Learning
Paper • 2511.20549 • Published • 27
Camera Control in Video Generation
-
BulletTime: Decoupled Control of Time and Camera Pose for Video Generation
Paper • 2512.05076 • Published • 8 -
DualCamCtrl: Dual-Branch Diffusion Model for Geometry-Aware Camera-Controlled Video Generation
Paper • 2511.23127 • Published • 44 -
Wan-Move: Motion-controllable Video Generation via Latent Trajectory Guidance
Paper • 2512.08765 • Published • 134
Diffusion Models
Camera Control in Video Generation
-
BulletTime: Decoupled Control of Time and Camera Pose for Video Generation
Paper • 2512.05076 • Published • 8 -
DualCamCtrl: Dual-Branch Diffusion Model for Geometry-Aware Camera-Controlled Video Generation
Paper • 2511.23127 • Published • 44 -
Wan-Move: Motion-controllable Video Generation via Latent Trajectory Guidance
Paper • 2512.08765 • Published • 134
AR Video Generation
-
Deep Forcing: Training-Free Long Video Generation with Deep Sink and Participative Compression
Paper • 2512.05081 • Published • 33 -
Reward Forcing: Efficient Streaming Video Generation with Rewarded Distribution Matching Distillation
Paper • 2512.04678 • Published • 42 -
What about gravity in video generation? Post-Training Newton's Laws with Verifiable Rewards
Paper • 2512.00425 • Published • 53 -
Flash-DMD: Towards High-Fidelity Few-Step Image Generation with Efficient Distillation and Joint Reinforcement Learning
Paper • 2511.20549 • Published • 27