DexGraspNet 2.0: Learning Generative Dexterous Grasping in Large-scale Synthetic Cluttered Scenes
Not on the map. Why?
Abstract (arXiv)
Grasping in cluttered scenes remains highly challenging for dexterous hands due to the scarcity of data. To address this problem, we present a large-scale synthetic benchmark, encompassing 1319 objects, 8270 scenes, and 427 million grasps. Beyond benchmarking, we also propose a novel two-stage grasping method that learns efficiently from data by using a diffusion model that conditions on local geometry. Our proposed generative method outperforms all baselines in simulation experiments. Furthermore, with the aid of test-time-depth restoration, our method demonstrates zero-shot sim-to-real transfer, attaining 90.7% real-world dexterous grasping success rate in cluttered scenes.
Key facts. Cited by 41 papers in the map (data through 7 Oct 2026). Builds on 35 of its 49 references that are in the map. First posted on arXiv 30 Oct 2024.
Builds on
35 of 49 references.
References · 35
- 2020Denoising Diffusion Probabilistic Modelscited by 20,851
- 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 8,964
- 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
- 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,384
- 2018Neural Ordinary Differential Equationscited by 3,733
- 2015ShapeNet: An Information-Rich 3D Model Repositorycited by 2,880
- 20194D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networkscited by 1,201
- 2021Isaac Gym: High Performance GPU-Based Physics Simulation For Robot Learningcited by 882
- 20196-DOF GraspNet: Variational Grasp Generation for Object Manipulationcited by 237
- 2021Contact-GraspNet: Efficient 6-DoF Grasp Generation in Cluttered Scenescited by 213
- 2022AnyGrasp: Robust and Efficient Grasp Perception in Spatial and Temporal Domainscited by 205
- 2021Hand-Object Contact Consistency Reasoning for Human Grasps Generationcited by 160
23 more
- 2022DexGraspNet: A Large-Scale Robotic Dexterous Grasp Dataset for General Objects Based on Simulationcited by 129
- 2018Closing the Loop for Robotic Grasping: A Real-time, Generative Grasp Synthesis Approachcited by 117
- 2023UniDexGrasp: Universal Robotic Dexterous Grasping via Learning Diverse Proposal Generation and Goal-Conditioned Policycited by 117
- 2020ACRONYM: A Large-Scale Grasp Dataset Based on Simulationcited by 115
- 2021D-Grasp: Physically Plausible Dynamic Grasp Synthesis for Hand-Object Interactionscited by 109
- 2017Grasp Pose Detection in Point Cloudscited by 107
- 2021Synergies Between Affordance and Geometry: 6-DoF Grasp Detection via Implicit Representationscited by 107
- 2023UniDexGrasp++: Improving Dexterous Grasping Policy Learning via Geometry-aware Curriculum and Iterative Generalist-Specialist Learningcited by 101
- 2021Volumetric Grasping Network: Real-time 6 DOF Grasp Detection in Cluttercited by 94
- 2020An Analysis of SVD for Deep Rotation Estimationcited by 85
- 2018PointNetGPD: Detecting Grasp Configurations from Point Setscited by 80
- 2023LEAP Hand: Low-Cost, Efficient, and Anthropomorphic Hand for Robot Learningcited by 78
- 2021Synthesizing Diverse and Physically Stable Grasps with Arbitrary Hand Structures using Differentiable Force Closure Estimatorcited by 70
- 2022Domain Randomization-Enhanced Depth Simulation and Restoration for Perceiving and Grasping Specular and Transparent Objectscited by 49
- 2024Graspness Discovery in Clutters for Fast and Accurate Grasp Detectioncited by 47
- 2023cuRobo: Parallelized Collision-Free Minimum-Jerk Robot Motion Generationcited by 46
- 2020Deep Differentiable Grasp Planner for High-DOF Gripperscited by 39
- 2023Dexterous Functional Graspingcited by 36
- 2021DDGC: Generative Deep Dexterous Grasping in Cluttercited by 28
- 2022Learning Robust Real-World Dexterous Grasping Policies via Implicit Shape Augmentationcited by 21
- 2018Multi-View Picking: Next-best-view Reaching for Improved Grasping in Cluttercited by 15
- 2024ContactDexNet: Multi-fingered Robotic Hand Grasping in Cluttered Environments through Hand-object Contact Semantic Mappingcited by 8
- 2019Increasing the Generalisation Capacity of Conditional VAEscited by 1
Built on by
Citing papers per year: 2024: 2; 2025: 15; 2026: 24.
41 papers.
Sort:
By area:
- 2025DexGraspVLA: A Vision-Language-Action Framework Towards General Dexterous Graspingcited by 87
- 2024Code-as-Monitor: Constraint-aware Visual Programming for Reactive and Proactive Robotic Failure Detectioncited by 59
- 2025Diffusion Models for Robotic Manipulation: A Surveycited by 39
- 2025Dex1B: Learning with 1B Demonstrations for Dexterous Manipulationcited by 23
- 2025Dexonomy: Synthesizing All Dexterous Grasp Types in a Grasp Taxonomycited by 19
- 2025DexVLG: Dexterous Vision-Language-Grasp Model at Scalecited by 19
- 2025Object-centric 3D Motion Field for Robot Learning from Human Videoscited by 13
- 2024Multi-GraspLLM: A Multimodal LLM for Multi-Hand Semantic Guided Grasp Generationcited by 12
- 2025EC-Flow: Enabling Versatile Robotic Manipulation from Action-Unlabeled Videos via Embodiment-Centric Flowcited by 10
- 2026Dex4D: Task-Agnostic Point Track Policy for Sim-to-Real Dexterous Manipulationcited by 10
- 2026SimToolReal: An Object-Centric Policy for Zero-Shot Dexterous Tool Manipulationcited by 10
- 2026Dexora: Open-source VLA for High-DoF Bimanual Dexteritycited by 9
29 more
- 2026DextER: Language-driven Dexterous Grasp Generation with Embodied Reasoningcited by 4
- 2026Data Pyramid for Embodied Manipulation: A Surveycited by 4
- 2025Joint Flow Trajectory Optimization For Feasible Robot Motion Generation from Video Demonstrationscited by 2
- 2025Lightning Grasp: High Performance Procedural Grasp Synthesis with Contact Fieldscited by 2
- 2026CoLA-Flow Policy: Temporally Coherent Imitation Learning via Continuous Latent Action Flow Matching for Robotic Manipulationcited by 2
- 2026VTouch++: A Multimodal Dataset with Vision-Based Tactile Enhancement for Bimanual Manipulationcited by 2
- 2026Bench2Dex: Benchmarking Visuo-Tactile Bimanual Dexterous Manipulation Across Dexterous Handscited by 2
- 2025Simultaneous Pick and Place Detection by Combining SE(3) Diffusion Models with Differential Kinematicscited by 1
- 2025Rodrigues Network for Learning Robot Actionscited by 1
- 2025Equivariant Volumetric Graspingcited by 1
- 2026D-REX: Differentiable Real-to-Sim-to-Real Engine for Learning Dexterous Graspingcited by 1
- 2026SIM1: Physics-Aligned Simulator as Zero-Shot Data Scaler in Deformable Worldscited by 1
- 2026X-DiffVLA: X-Embodied Diffusion Action Heads for Vision-Language-Action Modelscited by 1
- 2026EquiDexFlow: Contact-Grounded SE(3)-Equivariant Dexterous Grasp Generative Flowscited by 1
- 2026EAGG: Embodiment-Aligned Grasp Generation via Geometry-Aware Graph Conditioningcited by 1
- 2026AutoDex: An Automated Real-World System for Dexterous Grasping Data Collectioncited by 1
- 2026Play2Perfect: What Matters in Dexterous Play Pretraining for Precise Assembly?cited by 1
- 2026Adaptive Vision-Language Grasping via Composable Foundation Priors and Generalizable Grasp Synthesiscited by 1
- 2025LensDFF: Language-enhanced Sparse Feature Distillation for Efficient Few-Shot Dexterous Manipulationcited by 0
- 2025DCIRNet: Depth Completion with Iterative Refinement for Dexterous Grasping of Transparent and Reflective Objectscited by 0
- 2025DexCanvas: Bridging Human Demonstrations and Robot Learning for Dexterous Manipulationcited by 0
- 2026SECOND-Grasp: Semantic Contact-guided Dexterous Graspingcited by 0
- 2026DexHoldem: Playing Texas Hold'em with Dexterous Embodied Systemcited by 0
- 2026Simulation-Driven Imitation Learning for Biosignals-Free Shared-Autonomy Prosthetic Graspingcited by 0
- 2026KPGrasp: Scalable Keypoint Flow Matching for Dexterous Grasp Generationcited by 0
- 2026UniDexTok: A Unified Dexterous Hand Tokenizer from Real Datacited by 0
- 2026V2P-Manip: Learning Dexterous Manipulation from Monocular Human Videoscited by 0
- 2026DexCompose: Reusing Dexterous Policies for Multi-Task Manipulation with a Single Handcited by 0
- 2026UniPart: Towards Zero-shot Language-Grounded 3D Part Segmentation for Embodied Interactioncited by 0
- 2026Bench2Dex: Benchmarking Visuo-Tactile Bimanual Dexterous Manipulation Across Dexterous Hands14 Sep 2026
- 2026UniPart: Towards Zero-shot Language-Grounded 3D Part Segmentation for Embodied Interaction11 Sep 2026
- 2026Adaptive Vision-Language Grasping via Composable Foundation Priors and Generalizable Grasp Synthesis3 Sep 2026
- 2026Data Pyramid for Embodied Manipulation: A Survey27 Jul 2026
- 2026DexCompose: Reusing Dexterous Policies for Multi-Task Manipulation with a Single Hand26 Jun 2026
- 2026Play2Perfect: What Matters in Dexterous Play Pretraining for Precise Assembly?24 Jun 2026
- 2026AutoDex: An Automated Real-World System for Dexterous Grasping Data Collection22 Jun 2026
- 2026EAGG: Embodiment-Aligned Grasp Generation via Geometry-Aware Graph Conditioning16 Jun 2026
- 2026V2P-Manip: Learning Dexterous Manipulation from Monocular Human Videos15 Jun 2026
- 2026EquiDexFlow: Contact-Grounded SE(3)-Equivariant Dexterous Grasp Generative Flows10 Jun 2026
- 2026UniDexTok: A Unified Dexterous Hand Tokenizer from Real Data9 Jun 2026
- 2026KPGrasp: Scalable Keypoint Flow Matching for Dexterous Grasp Generation8 Jun 2026
29 more
- 2026Simulation-Driven Imitation Learning for Biosignals-Free Shared-Autonomy Prosthetic Grasping5 Jun 2026
- 2026X-DiffVLA: X-Embodied Diffusion Action Heads for Vision-Language-Action Models24 May 2026
- 2026DexHoldem: Playing Texas Hold'em with Dexterous Embodied System18 May 2026
- 2026Dexora: Open-source VLA for High-DoF Bimanual Dexterity18 May 2026
- 2026SECOND-Grasp: Semantic Contact-guided Dexterous Grasping13 May 2026
- 2026VTouch++: A Multimodal Dataset with Vision-Based Tactile Enhancement for Bimanual Manipulation22 Apr 2026
- 2026SIM1: Physics-Aligned Simulator as Zero-Shot Data Scaler in Deformable Worlds9 Apr 2026
- 2026D-REX: Differentiable Real-to-Sim-to-Real Engine for Learning Dexterous Grasping1 Mar 2026
- 2026SimToolReal: An Object-Centric Policy for Zero-Shot Dexterous Tool Manipulation18 Feb 2026
- 2026Dex4D: Task-Agnostic Point Track Policy for Sim-to-Real Dexterous Manipulation17 Feb 2026
- 2026CoLA-Flow Policy: Temporally Coherent Imitation Learning via Continuous Latent Action Flow Matching for Robotic Manipulation30 Jan 2026
- 2026DextER: Language-driven Dexterous Grasp Generation with Embodied Reasoning22 Jan 2026
- 2025Lightning Grasp: High Performance Procedural Grasp Synthesis with Contact Fields10 Nov 2025
- 2025DexCanvas: Bridging Human Demonstrations and Robot Learning for Dexterous Manipulation17 Oct 2025
- 2025Joint Flow Trajectory Optimization For Feasible Robot Motion Generation from Video Demonstrations25 Sep 2025
- 2025Equivariant Volumetric Grasping24 Jul 2025
- 2025EC-Flow: Enabling Versatile Robotic Manipulation from Action-Unlabeled Videos via Embodiment-Centric Flow8 Jul 2025
- 2025DexVLG: Dexterous Vision-Language-Grasp Model at Scale3 Jul 2025
- 2025Dex1B: Learning with 1B Demonstrations for Dexterous Manipulation20 Jun 2025
- 2025DCIRNet: Depth Completion with Iterative Refinement for Dexterous Grasping of Transparent and Reflective Objects11 Jun 2025
- 2025Object-centric 3D Motion Field for Robot Learning from Human Videos4 Jun 2025
- 2025Rodrigues Network for Learning Robot Actions3 Jun 2025
- 2025Simultaneous Pick and Place Detection by Combining SE(3) Diffusion Models with Differential Kinematics28 Apr 2025
- 2025Dexonomy: Synthesizing All Dexterous Grasp Types in a Grasp Taxonomy26 Apr 2025
- 2025Diffusion Models for Robotic Manipulation: A Survey11 Apr 2025
- 2025LensDFF: Language-enhanced Sparse Feature Distillation for Efficient Few-Shot Dexterous Manipulation5 Mar 2025
- 2025DexGraspVLA: A Vision-Language-Action Framework Towards General Dexterous Grasping28 Feb 2025
- 2024Multi-GraspLLM: A Multimodal LLM for Multi-Hand Semantic Guided Grasp Generation11 Dec 2024
- 2024Code-as-Monitor: Constraint-aware Visual Programming for Reactive and Proactive Robotic Failure Detection5 Dec 2024
- 2025Diffusion Models for Robotic Manipulation: A Surveycited by 39
- 2026CoLA-Flow Policy: Temporally Coherent Imitation Learning via Continuous Latent Action Flow Matching for Robotic Manipulationcited by 2
- 2025Simultaneous Pick and Place Detection by Combining SE(3) Diffusion Models with Differential Kinematicscited by 1
- 2026KPGrasp: Scalable Keypoint Flow Matching for Dexterous Grasp Generationcited by 0
- 2026KPGrasp: Scalable Keypoint Flow Matching for Dexterous Grasp Generation8 Jun 2026
- 2026CoLA-Flow Policy: Temporally Coherent Imitation Learning via Continuous Latent Action Flow Matching for Robotic Manipulation30 Jan 2026
- 2025Simultaneous Pick and Place Detection by Combining SE(3) Diffusion Models with Differential Kinematics28 Apr 2025
- 2025Diffusion Models for Robotic Manipulation: A Survey11 Apr 2025
- 2025DexGraspVLA: A Vision-Language-Action Framework Towards General Dexterous Graspingcited by 87
- 2025DexVLG: Dexterous Vision-Language-Grasp Model at Scalecited by 19
- 2026Dexora: Open-source VLA for High-DoF Bimanual Dexteritycited by 9
- 2026X-DiffVLA: X-Embodied Diffusion Action Heads for Vision-Language-Action Modelscited by 1
- 2026X-DiffVLA: X-Embodied Diffusion Action Heads for Vision-Language-Action Models24 May 2026
- 2026Dexora: Open-source VLA for High-DoF Bimanual Dexterity18 May 2026
- 2025DexVLG: Dexterous Vision-Language-Grasp Model at Scale3 Jul 2025
- 2025DexGraspVLA: A Vision-Language-Action Framework Towards General Dexterous Grasping28 Feb 2025
- 2024Code-as-Monitor: Constraint-aware Visual Programming for Reactive and Proactive Robotic Failure Detectioncited by 59
- 2025Dex1B: Learning with 1B Demonstrations for Dexterous Manipulationcited by 23
- 2025Dexonomy: Synthesizing All Dexterous Grasp Types in a Grasp Taxonomycited by 19
- 2025Object-centric 3D Motion Field for Robot Learning from Human Videoscited by 13
- 2024Multi-GraspLLM: A Multimodal LLM for Multi-Hand Semantic Guided Grasp Generationcited by 12
- 2026Dex4D: Task-Agnostic Point Track Policy for Sim-to-Real Dexterous Manipulationcited by 10
- 2026SimToolReal: An Object-Centric Policy for Zero-Shot Dexterous Tool Manipulationcited by 10
- 2026Data Pyramid for Embodied Manipulation: A Surveycited by 4
- 2025Joint Flow Trajectory Optimization For Feasible Robot Motion Generation from Video Demonstrationscited by 2
- 2025Lightning Grasp: High Performance Procedural Grasp Synthesis with Contact Fieldscited by 2
- 2026VTouch++: A Multimodal Dataset with Vision-Based Tactile Enhancement for Bimanual Manipulationcited by 2
- 2026Bench2Dex: Benchmarking Visuo-Tactile Bimanual Dexterous Manipulation Across Dexterous Handscited by 2
12 most cited of 29 All 29 (JSON)
- 2026Bench2Dex: Benchmarking Visuo-Tactile Bimanual Dexterous Manipulation Across Dexterous Hands14 Sep 2026
- 2026UniPart: Towards Zero-shot Language-Grounded 3D Part Segmentation for Embodied Interaction11 Sep 2026
- 2026Adaptive Vision-Language Grasping via Composable Foundation Priors and Generalizable Grasp Synthesis3 Sep 2026
- 2026Data Pyramid for Embodied Manipulation: A Survey27 Jul 2026
- 2026DexCompose: Reusing Dexterous Policies for Multi-Task Manipulation with a Single Hand26 Jun 2026
- 2026Play2Perfect: What Matters in Dexterous Play Pretraining for Precise Assembly?24 Jun 2026
- 2026AutoDex: An Automated Real-World System for Dexterous Grasping Data Collection22 Jun 2026
- 2026EAGG: Embodiment-Aligned Grasp Generation via Geometry-Aware Graph Conditioning16 Jun 2026
- 2026V2P-Manip: Learning Dexterous Manipulation from Monocular Human Videos15 Jun 2026
- 2026EquiDexFlow: Contact-Grounded SE(3)-Equivariant Dexterous Grasp Generative Flows10 Jun 2026
- 2026UniDexTok: A Unified Dexterous Hand Tokenizer from Real Data9 Jun 2026
- 2026SECOND-Grasp: Semantic Contact-guided Dexterous Grasping13 May 2026
12 newest of 29 All 29 (JSON)
Similar, not cited
Close in wording, no citation either way
- 2025ClutterDexGrasp: A Sim-to-Real System for General Dexterous Grasping in Cluttered Scenes
- 2026CADGrasp: Learning Contact and Collision Aware General Dexterous Grasping in Cluttered Scenes
- 2025DexGrasp Anything: Towards Universal Robotic Dexterous Grasping with Physics Awareness
- 2024DexGANGrasp: Dexterous Generative Adversarial Grasping Synthesis for Task-Oriented Manipulation
- 2024DexDiffuser: Generating Dexterous Grasps with Diffusion Models
- 2023GraspLDM: Generative 6-DoF Grasp Synthesis using Latent Diffusion Models
- 2025G-DexGrasp: Generalizable Dexterous Grasping Synthesis Via Part-Aware Prior Retrieval and Prior-Assisted Generation
- 2024DexGrasp-Diffusion: Diffusion-based Unified Functional Grasp Synthesis Method for Multi-Dexterous Robotic Hands