ResVLA: From Noise to Intent: Anchoring Generative VLA Policies with Residual Bridges

Yiming Zhong*, Yaoyu He*, Zemin Yang*, Pengfei Tian, Yifan Huang, Qingqiu Huang, Xinge Zhu, Yuexin Ma†

This paper proposes ResVLA, a generative vision-language-action framework that shifts robot control from generation-from-noise to refinement-from-intent by anchoring low-frequency semantic intent and refining high-frequency residual dynamics, achieving strong robustness, faster convergence, and competitive real-world performance.

Affordance-R1: Reinforcement Learning for Generalizable Affordance Reasoning in Multimodal Large Language Model

Hanqing Wang*, Shaoyang Wang*, Yiming Zhong, Zemin Yang, Jiamin Wang, Zhiqing Cui, Jiahao Yuan, Yifan Han, Mingyu Liu, Yuexin Ma†

We introduce Affordance-R1, which is capable of generating explicit reasoning alongside the final answer. With the help of proposed affordance reasoning reward, it achieves robust zero-shot generalization and exhibits emergent test-time reasoning capabilities.

HUMOF: Human Motion Forecasting in Interactive Social Scenes

Caiyi Sun*, Yujing Sun*, Xiao Han, Zemin Yang, Jiawei Liu, Xinge Zhu, Siu-Ming Yiu†, Yuexin Ma†

This paper proposes HUMOF, a human motion forecasting framework for interactive social scenes that jointly models human-human and human-scene interactions with hierarchical representations and coarse-to-fine reasoning, achieving state-of-the-art performance across four public datasets.

FreqPolicy: Frequency Autoregressive Visuomotor Policy with Continuous Tokens

Yiming Zhong, Yumeng Liu, Chuyang Xiao, Zemin Yang, Youzhuo Wang, Yufei Zhu, Yujing Sun, Xinge Zhu, Yuexin Ma†

This paper proposes FreqPolicy, a frequency-domain autoregressive visuomotor policy that progressively models hierarchical frequency components with continuous latent representations, achieving superior accuracy and efficiency in robotic manipulation tasks.

EvolvingGrasp: Evolutionary Grasp Generation via Efficient Preference Alignment

Yufei Zhu*, Yiming Zhong*, Zemin Yang, Peishan Cong, Jingyi Yu, Xinge Zhu†, Yuexin Ma†

This paper introduces EvolvingGrasp, which integrates Handpose-wise Preference Optimization with a Physics-aware Consistency Model to enable efficient evolutionary grasp generation, achieving improved grasp success rates and computational efficiency.

EasyHOI: Unleashing the Power of Large Models for Reconstructing Hand-Object Interactions in the Wild

Yumeng Liu, Xiaoxiao Long†, Zemin Yang, Yuan Liu, Marc Habermann, Christian Theobalt, Yuexin Ma†, Wenping Wang

This paper explores reconstructing hand-object interactions from a single-view image. We develop a pipeline to estimate hand pose and object shape using large models and apply a prior-guided optimization to adjust the hand pose, ensuring it meets 3D physical constraints while aligning with the 2D image.

UniDemoiré: Towards Universal Image Demoiréing with Data Generation and Synthesis

Zemin Yang*, Yujing Sun*, Xidong Peng, Siu Ming Yiu, Yuexin Ma†

This paper proposes a universal image demoiréing solution, UniDemoiré, which has superior generalization capability. Notably, we propose innovative and effective data generation and synthesis methods that can automatically provide vast high-quality moiré images to train a universal demoiréing model.