4 papers
Dynamics-Aware Meta-Imitation for Generalization to Unseen Robotic Manipulation
Zhenduo Shang, Xiyao Liu, Bohan Li +4
Imitation Learning aims to learn skills from extensive observations and demonstrations for robots, so it suffers from data scarcity and environment generalization. The existing met…
Seeing Before Reasoning: Decoupling Perception and Reasoning for Shortcut-Resilient Multimodal On-Policy Self-Distillation
Sihan Wang, Xiyao Liu, Lianqing Liu +1
On-policy self-distillation (OPSD) trains a model on its own rollouts and uses a frozen copy to provide dense token-level targets conditioned on a reference target. This works well…
Vision and Language Integration for Domain Generalization
Yanmei Wang, Xiyao Liu, Fupeng Chu +1
Domain generalization aims at training on source domains to uncover a domain-invariant feature space, allowing the model to perform robust generalization ability on unknown target…
Unbiased Faster R-CNN for Single-source Domain Generalized Object Detection
Yajing Liu, Shijun Zhou, Xiyao Liu +3
Single-source domain generalization (SDG) for object detection is a challenging yet essential task as the distribution bias of the unseen domain degrades the algorithm performance…