15 papers
WoVR+: World Model as Reliable Simulator for Versatile VLA Policy Optimization with RL
Shangqing Zhou, Zhennan Jiang, Ning Ma +9
Action-conditioned world models can reduce costly real-robot interaction in reinforcement learning for Vision-Language-Action models (VLA-RL), but hallucinations can mislead policy…
Harnessing VLM in Embodied RL: From Reward Regressor to Success Detector via Exploration-Augmented Key-Observation Gating
Xin Liu, Yutong Jiang, Jian Su +4
VLMs provide promising solutions to embodied reward design. Most current methods that train VLMs as reward regressors not only require extensive annotations but are also primarily…
GIFT: Guided Intermediate Feature Training via Action-Oriented Structural Supervision for Robotic Manipulation
Yupeng Zheng, Xiang Li, Songen Gu +11
Vision-language pre-training and predictive world modeling provide robot policies with rich semantic and dynamic visual features, but their native action and visual-prediction obje…
WALA Learning Executable Latent Actions from Action-Labeled Demonstrations and Action-Free Videos
Jiahao Liu, Zhongpu Xia, Yupeng Zheng +14
Human videos provide rich information about object manipulation at a scale difficult to reproduce with robots, but often lack action annotations that can directly supervise robot p…
VT-WAM: Visual-Tactile World Action Model for Contact-Rich Manipulation
Shuai Tian, Yupeng Zheng, Yuhang Zheng +7
Contact-rich manipulation requires policies to react to local deformation, pressure, slip, and friction, yet these cues are temporally sparse and often invisible in visual observat…
Dual-Granularity Contrastive Reward via Generated Episodic Guidance for Efficient Embodied RL
Xin Liu, Yixuan Li, Yuhui Chen +3
Designing suitable rewards poses a significant challenge in reinforcement learning (RL), especially for embodied manipulation. Trajectory success rewards are suitable for human jud…
WoVR: World Models as Reliable Simulators for Post-Training VLA Policies with RL
Zhennan Jiang, Shangqing Zhou, Yutong Jiang +11
Reinforcement learning (RL) promises to unlock capabilities beyond imitation learning for Vision--Language--Action (VLA) models, but its requirement for massive real-world interact…
Towards Long-Lived Robots: Continual Learning VLA Models via Reinforcement Fine-Tuning
Yuan Liu, Haoran Li, Shuai Tian +5
Pretrained on large-scale and diverse datasets, VLA models demonstrate strong generalization and adaptability as general-purpose robotic policies. However, Supervised Fine-Tuning (…
DiffuDepGrasp: Diffusion-based Depth Noise Modeling Empowers Sim2Real Robotic Grasping
Yingting Zhou, Wenbo Cui, Weiheng Liu +3
Transferring the depth-based end-to-end policy trained in simulation to physical robots can yield an efficient and robust grasping policy, yet sensor artifacts in real depth maps l…
QDepth-VLA: Quantized Depth Prediction as Auxiliary Supervision for Vision-Language-Action Models
Yixuan Li, Yuhui Chen, Mingcai Zhou +3
Spatial perception and reasoning are crucial for Vision-Language-Action (VLA) models to accomplish fine-grained manipulation tasks. However, existing approaches often lack the abil…
World4RL: Diffusion World Models for Policy Refinement with Reinforcement Learning for Robotic Manipulation
Zhennan Jiang, Kai Liu, Yuxin Qin +6
Robotic manipulation policies are commonly initialized through imitation learning, but their performance is limited by the scarcity and narrow coverage of expert data. Reinforcemen…
Survey of Vision-Language-Action Models for Embodied Manipulation
Haoran Li, Yuhui Chen, Wenbo Cui +5
Embodied intelligence systems, which enhance agent capabilities through continuous environment interactions, have garnered significant attention from both academia and industry. Vi…
CLAR: Learning 3D Representations for Robotic Manipulation by Fusing Masked Reconstruction with Multi-Level Contrastive Alignment
Wenbo Cui, Chengyang Zhao, Yuhui Chen +4
The spatial information inherent in 3D point clouds is crucial for robotic manipulation. However, existing 3D pre-training methods face a fundamental trade-off: Masked Autoencoding…
GMatch: A Lightweight, Geometry-Constrained Keypoint Matcher for Zero-Shot 6DoF Pose Estimation in Robotic Grasp Tasks
Ming Yang, Haoran Li
6DoF object pose estimation is fundamental to robotic grasp tasks. While recent learning-based methods achieve high accuracy, their computational demands hinder deployment on resou…
FetchBot: Learning Generalizable Object Fetching in Cluttered Scenes via Zero-Shot Sim2Real
Weiheng Liu, Yuxuan Wan, Jilong Wang +7
Generalizable object fetching in cluttered scenes remains a fundamental and application-critical challenge in embodied AI. Closely packed objects cause inevitable occlusions, makin…