VLA [CMU|Pathak]
ViPRA: Video Prediction for Robot Actions
Sandeep Routray et al. · Video prediction to policy conversion, unlabeled video usage.
cs.RO
閱讀原文
VLA 研究日報 Pulsar
LIVE
— AI 線今日無資料 —— AI 线今日无资料 —
VLA 線VLA 线 · cs.RO · cs.AI · cs.LG
- ViPRA: Video Prediction for Robot Actions [CMU|Pathak] Sandeep Routray et al. · Video prediction to policy conversion, unlabeled video usage. CS.RO
- EgoDemoGen: Egocentric Demonstration Generation for Viewpoint Generalization in Robotic Manipulation [清华] Yuan Xu et al. · Egocentric viewpoint augmentation for IL robustness. CS.RO
- RoboManipBaselines: A Unified Framework for Imitation Learning in Robotic Manipulation across Real and Simulation Environments Masaki Murooka et al. · Open-source IL framework for real/sim pipeline. CS.RO
- Goal-VLA: Image-Generative VLMs as Object-Centric World Models Empowering Zero-shot Robot Manipulation [清华] Haonan Chen et al. · Generative VLM as world model for zero-shot manipulation. CS.RO
- 3D CAVLA: Leveraging Depth and 3D Context to Generalize Vision Language Action Models for Unseen Tasks Vineet Bhat et al. · Integrates depth/3D context into VLA for task generalization. CS.RO
- Continual Robot Skill and Task Learning via Dialogue Weiwei Gu et al. · Dialogue-based continual skill learning, interaction focus. CS.RO
- FocusVLA: Focused Visual Utilization for Vision-Language-Action Models Yichi Zhang et al. · VLA attention improvement, overlaps with tracked Selective Perception [💧灌水]. CS.RO
- SOLE-R1: Video-Language Reasoning as the Sole Reward for On-Robot Reinforcement Learning Philip Schroeder et al. · Uses VLM reasoning as reward signal for on-robot RL. CS.RO
- StreamingVLA: Streaming Vision-Language-Action Model with Action Flow Matching and Adaptive Early Observation [浙大] Yiran Shi et al. · Overlaps with tracked StreamVLA, efficiency optimization [💧灌水]. CS.RO
- ManipArena: Comprehensive Real-world Evaluation of Reasoning-Oriented Generalist Robot Manipulation [清华] Yu Sun et al. · New benchmark for reasoning-oriented manipulation evaluation. CS.RO
- Feel Robot Feels: Tactile Feedback Array Glove for Dexterous Manipulation [清华] Feiyu Jia et al. · Tactile glove hardware for dexterous teleoperation data collection. CS.RO
- Tac2Real: Reliable and GPU Visuotactile Simulation for Online Reinforcement Learning and Zero-Shot Real-World Deployment [科大] Ningyu Yan et al. · GPU-accelerated visuotactile sim for online RL and zero-shot deployment. CS.RO
VLA 研究日報VLA 研究日报
共 28 篇
🔧 技術技术
VLA [清华]
Goal-VLA: Image-Generative VLMs as Object-Centric World Models Empowering Zero-shot Robot Manipulation
Haonan Chen et al. · Generative VLM as world model for zero-shot manipulation.
cs.RO
閱讀原文
VLA [科大]
Tac2Real: Reliable and GPU Visuotactile Simulation for Online Reinforcement Learning and Zero-Shot Real-World Deployment
Ningyu Yan et al. · GPU-accelerated visuotactile sim for online RL and zero-shot deployment.
cs.RO
閱讀原文
VLA [北大|Mu]
ProgressVLA: Progress-Guided Diffusion Policy for Vision-Language Robotic Manipulation
Hongyu Yan et al. · Progress guidance for VLA diffusion policy, solves long-horizon termination.
cs.RO
閱讀原文
VLA
HiFlow: Tokenization-Free Scale-Wise Autoregressive Policy Learning via Flow Matching
Daichi Yashima et al. · Tokenization-free flow matching policy, efficient autoregressive learning.
cs.RO
閱讀原文
📖 背景閱讀背景阅读
VLA [清华]
EgoDemoGen: Egocentric Demonstration Generation for Viewpoint Generalization in Robotic Manipulation
Yuan Xu et al. · Egocentric viewpoint augmentation for IL robustness.
cs.RO
閱讀原文
VLA
RoboManipBaselines: A Unified Framework for Imitation Learning in Robotic Manipulation across Real and Simulation Environments
Masaki Murooka et al. · Open-source IL framework for real/sim pipeline.
cs.RO
閱讀原文
VLA
3D CAVLA: Leveraging Depth and 3D Context to Generalize Vision Language Action Models for Unseen Tasks
Vineet Bhat et al. · Integrates depth/3D context into VLA for task generalization.
cs.RO
閱讀原文
VLA
Continual Robot Skill and Task Learning via Dialogue
Weiwei Gu et al. · Dialogue-based continual skill learning, interaction focus.
cs.RO
閱讀原文
VLA
FocusVLA: Focused Visual Utilization for Vision-Language-Action Models
Yichi Zhang et al. · VLA attention improvement, overlaps with tracked Selective Perception [💧灌水].
cs.RO
閱讀原文
VLA
SOLE-R1: Video-Language Reasoning as the Sole Reward for On-Robot Reinforcement Learning
Philip Schroeder et al. · Uses VLM reasoning as reward signal for on-robot RL.
cs.RO
閱讀原文
VLA [浙大]
StreamingVLA: Streaming Vision-Language-Action Model with Action Flow Matching and Adaptive Early Observation
Yiran Shi et al. · Overlaps with tracked StreamVLA, efficiency optimization [💧灌水].
cs.RO
閱讀原文
VLA [清华]
ManipArena: Comprehensive Real-world Evaluation of Reasoning-Oriented Generalist Robot Manipulation
Yu Sun et al. · New benchmark for reasoning-oriented manipulation evaluation.
cs.RO
閱讀原文
VLA [清华]
Feel Robot Feels: Tactile Feedback Array Glove for Dexterous Manipulation
Feiyu Jia et al. · Tactile glove hardware for dexterous teleoperation data collection.
cs.RO
閱讀原文
VLA
Tele-Catch: Adaptive Teleoperation for Dexterous Dynamic 3D Object Catching
Weiguang Zhao et al. · Teleoperation framework for dynamic object catching tasks.
cs.RO
閱讀原文
VLA [Unitree]
Active Stereo-Camera Outperforms Multi-Sensor Setup in ACT Imitation Learning for Humanoid Manipulation
Robin Kühn et al. · Empirical study on sensor setups for ACT on humanoids.
cs.RO
閱讀原文
VLA [MIT|Kaelbling]
Which Reconstruction Model Should a Robot Use? Routing Image-to-3D Models for Cost-Aware Robotic Manipulation
Akash Anand et al. · Routing mechanism for selecting 3D reconstruction models based on cost.
cs.RO
閱讀原文
VLA
Spectral Decomposition of Inverse Dynamics for Fast Exploration in Model-Based Manipulation
Solvin Sigurdson et al. · Spectral decomposition for inverse dynamics in model-based control.
cs.RO
閱讀原文
VLA [中科院]
Learning Smooth and Robust Space Robotic Manipulation of Dynamic Target via Inter-frame Correlation
Siyi Lang et al. · Control policy for space robotic manipulation of dynamic targets.
cs.RO
閱讀原文
VLA
D-SPEAR: Dual-Stream Prioritized Experience Adaptive Replay for Stable Reinforcement Learninging Robotic Manipulation
Yu Zhang et al. · Experience replay modification for stable RL in manipulation.
cs.RO
閱讀原文
VLA
UMI-Underwater: Learning Underwater Manipulation without Underwater Teleoperation
Hao Li et al. · Extends UMI framework to underwater manipulation without teleop.
cs.RO
閱讀原文
VLA
ROSClaw: An OpenClaw ROS 2 Framework for Agentic Robot Control and Interaction
Irvin Steve Cardenas et al. · ROS 2 framework for agentic robot control integration.
cs.RO
閱讀原文
VLA [浙大]
Beyond Viewpoint Generalization: What Multi-View Demonstrations Offer and How to Synthesize Them for Robot Manipulation?
Boyang Cai et al. · Systematic study on multi-view demonstration benefits and synthesis.
cs.RO
閱讀原文
VLA
Why Cognitive Robotics Matters: Lessons from OntoAgent and LLM Deployment in HARMONIC for Safety-Critical Robot Teaming
Sanjay Oruganti et al. · Cognitive architecture for safety-critical robot teaming with LLMs.
cs.RO
閱讀原文
VLA
Surface-Constrained Offline Warping with Contact-Aware Online Pose Projection for Safe Robotic Trajectory Execution
Farong Wang et al. · Safe trajectory execution via surface-constrained warping and pose projection.
cs.RO
閱讀原文
VLA [MIT]
Contextual Graph Representations for Task-Driven 3D Perception and Planning
Christopher Agia · Uses 3D scene graphs for task-driven perception and planning.
cs.RO
閱讀原文
VLA
Reducing Oracle Feedback with Vision-Language Embeddings for Preference-Based RL
Uses VLE to reduce oracle feedback cost in preference RL.
hf-papers
閱讀原文
VLA [LIBERO Team]
LIBERO-Para: A Diagnostic Benchmark and Metrics for Paraphrase Robustness in VLA Models
Benchmark for paraphrase robustness in VLA models on LIBERO.
hf-papers
閱讀原文