3 papers
cs.RO2026
Data-Free Continual Learning for World Action Model with Rectified Generative Replay
Kaipeng Fang, Hangxu Liu, Chaojian Shi +5
World action models (WAMs) built on pretrained video generation models adapt quickly to new robotic tasks, but they forget previously learned tasks under sequential fine-tuning. Ex…
cs.CV2025
Decoupled Diffusion Sparks Adaptive Scene Generation
Yunsong Zhou, Naisheng Ye, William Ljungbergh +6
Controllable scene generation could reduce the cost of diverse data collection substantially for autonomous driving. Prior works formulate the traffic layout generation as predicti…
cs.CV2025
Detect Anything 3D in the Wild
Hanxue Zhang, Haoran Jiang, Qingsong Yao +6
Despite the success of deep learning in close-set 3D object detection, existing approaches struggle with zero-shot generalization to novel objects and camera configurations. We int…