8 papers
OpenWAM: An Open, Modular Exploration Towards Systematic World-Action Model Pretraining
Yuran Wang, Siqiao Huang, Mingleyang Li +21
World-Action Models inherit world knowledge from video-generative priors, and channel it into executable control signals through embodied experience. Existing systems, however, are…
Scaling Bimanual Household Manipulation from 1,500 hours of Demonstrations to On-Policy Corrections
Jiafeng Xu, Qi Li, Yan Shen +7
Learning generalist policies for robust bimanual manipulation is bottlenecked by the scarcity of high quality large scale human demonstration data. In this work, we release 1,500 h…
EagleVLA: Towards Onboard Real-Time Robot Control via Foresight-Aligned Asynchronous Inference
Zebin Yang, Qi Wang, Yunhe Wang +6
Vision-Language-Action (VLA) models have achieved impressive performance on diverse embodied tasks. However, deploying VLA models on low-power onboard devices, such as the Jetson O…
AnchorVLA4D: an Anchor-Based Spatial-Temporal Vision-Language-Action Model for Robotic Manipulation
Juan Zhu, Zhanying Shao, Xiaoqi Li +4
Since current Vision-Language-Action (VLA) systems suffer from limited spatial perception and the absence of memory throughout manipulation, we investigate visual anchors as a mean…
Imagine2Act: Leveraging Object-Action Motion Consistency from Imagined Goals for Robotic Manipulation
Liang Heng, Jiadong Xu, Yiwen Wang +6
Relational object rearrangement (ROR) tasks (e.g., insert flower to vase) require a robot to manipulate objects with precise semantic and geometric reasoning. Existing approaches e…
RwoR: Generating Robot Demonstrations from Human Hand Collection for Policy Learning without Robot
Liang Heng, Xiaoqi Li, Shangqing Mao +9
Recent advancements in imitation learning have shown promising results in robotic manipulation, driven by the availability of high-quality training data. To improve data collection…
CrayonRobo: Object-Centric Prompt-Driven Vision-Language-Action Model for Robotic Manipulation
Xiaoqi Li, Lingyun Xu, Mingxu Zhang +8
In robotic, task goals can be conveyed through various modalities, such as language, goal images, and goal videos. However, natural language can be ambiguous, while images or video…
3DWG: 3D Weakly Supervised Visual Grounding via Category and Instance-Level Alignment
Xiaoqi Li, Jiaming Liu, Nuowei Han +4
The 3D weakly-supervised visual grounding task aims to localize oriented 3D boxes in point clouds based on natural language descriptions without requiring annotations to guide mode…