目前的一些工作：lqts.github.io
-
抓取
- UniDexGrasp: Universal Robotic Dexterous Grasping via Learning Diverse Proposal Generation and Goal-Conditioned Policy
- UniDexGrasp++: Improving Dexterous Grasping Policy Learning via Geometry-aware Curriculum and Iterative Generalist-Specialist Learning
- Text2HOI: Text-guided 3D Motion Generation for Hand-Object Interaction
- NL2Contact: Natural Language Guided 3D Hand-Object Contact Modeling with Diffusion Model
-
遥操作
- AnyTeleop: A General Vision-Based Dexterous Robot Arm-Hand Teleoperation System
-
预训练
- R3M: A Universal Visual Representation for Robot Manipulation
- Real-World Robot Learning with Masked Visual Pre-training
- The Unsurprising Effectiveness of Pre-Trained Vision Models for Control
- LIV: Language-Image Representations and Rewards for Robotic Control
- VIP: Towards Universal Visual Reward and Representation via Value-Implicit Pre-Training
- Language-Driven Representation Learning for Robotics
- Any-point Trajectory Modeling for Policy Learning
- Dexterity from Touch: Self-Supervised Pre-Training of Tactile Representations with Robotic Play
- Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point Modeling
-
灵巧操作
- DexArt: Benchmarking Generalizable Dexterous Manipulation With Articulated Objects
- A Multi-Agent Approach for Adaptive Finger Cooperation in Learning-based In-Hand Manipulation
- A Multi-Agent Approach for Adaptive Finger Cooperation in Learning-based In-Hand Manipulation
- DexPBT: Scaling up Dexterous Manipulation for Hand-Arm Systems with Population Based Training
- Toward Human-Like Grasp: Functional Grasp by Dexterous Robotic Hand Via Object-Hand Semantic Representation
- DeXtreme: Transfer of Agile In-hand Manipulation from Simulation to Reality
- Robot Synesthesia: In-Hand Manipulation with Visuotactile Sensing
- Towards Human-Level Bimanual Dexterous Manipulation with Reinforcement Learning
- A System for General In-Hand Object Re-Orientation
- General In-Hand Object Rotation with Vision and Touch
- Visual Dexterity: In-Hand Reorientation of Novel and Complex Object Shapes
- SimPLE, a visuotactile method learned in simulation to precisely pick, localize, regrasp, and place objects
-
长序列
- Hierarchical reinforcement learning for in-hand robotic manipulation using Davenport chained rotations
- Learning Hierarchical Control for Robust In-Hand Manipulation
- Sequential Dexterity: Chaining Dexterous Policies for Long-Horizon Manipulation
-
视触任务
- 3D Shape Reconstruction from Vision and Touch
- Active 3D Shape Reconstruction from Vision and Touch
- Visual-Tactile Sensing for In-Hand Object Reconstruction
- VisuoTactile 6D Pose Estimation of an In-Hand Object Using Vision and Tactile Sensor Data
- Hierarchical Graph Neural Networks for Proprioceptive 6D Pose Estimation of In-hand Objects
-
操作大模型
- Language-Image Goal-Conditioned Value Learning
- SayCan: Do As I Can, Not As I Say: Grounding Language in Robotic Affordances [Paper][Project][Code]
- Zero-Shot Reward Specification via Grounded Natural Language [Paper]
- VoxPoser: Composable 3D Value Maps for Robotic Manipulation with Language Models [Project]
- VIP: Towards Universal Visual Reward and Representation via Value-Implicit Pre-Training [Paper][Project]
- LIV: Language-Image Representations and Rewards for Robotic Control [Paper][Project]
- LOReL: Learning Language-Conditioned Robot Behavior from Offline Data and Crowd-Sourced Annotation [Paper][Project]
- Text2Motion: From Natural Language Instructions to Feasible Plans [Paper][Project]
- MineDreamer: Learning to Follow Instructions via Chain-of-Imagination for Simulated-World Control [Paper][Project][Code]
- Robot Transformers
- MotionGPT: Finetuned LLMs are General-Purpose Motion Generators [Paper][Project]
- RT-1: Robotics Transformer for Real-World Control at Scale [Paper][Project][Code]
- Masked Visual Pre-training for Motor Control [Paper][Project][Code]
- Real-world robot learning with masked visual pre-training [Paper][Project]
- R3M: A Universal Visual Representation for Robot Manipulation [Paper][Project][Code]
- Robot Learning with Sensorimotor Pre-training [Paper][Project]
- Rt-2: Vision-language-action models transfer web knowledge to robotic control [Paper][Project]
- PACT: Perception-Action Causal Transformer for Autoregressive Robotics Pre-Training [Paper]
- GROOT: Learning to Follow Instructions by Watching Gameplay Videos [Paper][Project][Code]
- Behavior Transformers (BeT): Cloning k modes with one stone [Paper][Project][Code]
- Conditional Behavior Transformers (C-BeT), From Play to Policy: Conditional Behavior Generation from Uncurated Robot Data [Paper][Project][Code]
- MAGIC: Meta-learning Adaptation for Ground Interaction Control with Visual Foundation Models [Paper]
- Language-Image Goal-Conditioned Value Learning
-
操作数据集
-
人手操作数据集
-
机器人操作数据集
MIME https://sites.google.com/view/mimedataset/dataset ROBOTURK https://arxiv.org/pdf/1811.02790 RoboNet https://arxiv.org/pdf/1910.11215 Bridge Data https://arxiv.org/pdf/2109.13396 BC-Z https://arxiv.org/pdf/2202.02005 RT-1 https://arxiv.org/pdf/2212.06817 RoboSet https://arxiv.org/pdf/2309.01918 Bridge Data v2 https://arxiv.org/pdf/2308.12952 RH20T https://arxiv.org/pdf/2307.00595 RealDex https://arxiv.org/pdf/2402.13853 DROID https://arxiv.org/abs/2403.12945
-