11 papers
Alleviating Sparse Rewards by Modeling Step-Wise and Long-Term Sampling Effects in Flow-Based GRPO
Yunze Tong, Mushui Liu, Canyu Zhao +7
Deploying GRPO on Flow Matching models has proven effective for text-to-image generation. However, existing paradigms typically propagate an outcome-based reward to all preceding d…
MMControl: Unified Multi-Modal Control for Joint Audio-Video Generation
Liyang Li, Wen Wang, Canyu Zhao +4
Recent advances in Diffusion Transformers (DiTs) have enabled high-quality joint audio-video generation, producing videos with synchronized audio within a single model. However, ex…
GAE: Unleashing Physical Potential of VLM with Generalizable Action Expert
Mingyu Liu, Zheng Huang, Xiaoyi Lin +6
Vision-language models demonstrate strong reasoning and planning abilities, yet grounding these predictions into precise robot actions remains a central challenge. Existing Vision-…
ACTIVE-o3: Empowering MLLMs with Active Perception via Pure Reinforcement Learning
Muzhi Zhu, Hao Zhong, Canyu Zhao +9
Active vision, also known as active perception, refers to actively selecting where and how to look in order to gather task-relevant information. It is a critical component of effic…
NoTVLA: Semantics-Preserving Robot Adaptation via Narrative Action Interfaces
Zheng Huang, Mingyu Liu, Xiaoyi Lin +9
Vision-Language-Action (VLA) models represent a pivotal advance in embodied intelligence, yet they confront critical barriers to real-world deployment, most notably catastrophic fo…
MARBLE: Multi-Aspect Reward Balance for Diffusion RL
Canyu Zhao, Hao Chen, Yunze Tong +3
Reinforcement learning fine-tuning has become the dominant approach for aligning diffusion models with human preferences. However, assessing images is intrinsically a multi-dimensi…