Carnegie Mellon University
LEAP Hand
A research hand with design files, APIs and examples for teleoperation and learned manipulation.
DEXTEROUS MANIPULATION
The world of dexterous hands.
Compare hardware. Discover applications. Share what works. An open reference and community for dexterous manipulation.
Projects
Explore source-linked robotics projects and tools on Dexhands.
UC San Diego / NVIDIA
AnyTeleop
Vision-based hand–arm teleoperation research with links to retargeting and visualization code.
Hugging Face / LeRobot contributors
LeRobot
Tools for recording datasets, training policies and exploring robot-learning workflows.
Google DeepMind / MuJoCo contributors
MuJoCo
A physics engine for articulated systems and contact-rich robotics experiments.
DexSuite contributors
Dex Retargeting
Software for mapping human hand motion to the joint configurations of robotic hands.
Meta research / TACTO contributors
TACTO
An archived research simulator for images from vision-based tactile sensors, including DIGIT and OmniTact examples.
Google DeepMind / Menagerie contributors
MuJoCo Menagerie
A collection of robot models for MuJoCo, including LEAP, Allegro, Shadow Hand and DEX-EE models.
SJTU / Tsinghua / UC San Diego research team
DexArt
A simulated benchmark for learning to manipulate articulated objects and testing generalization to unseen objects.
Google DeepMind / Playground contributors
MuJoCo Playground
GPU-accelerated robot-learning environments, including tasks for dexterous manipulation.
RLWRLD
DexBench industrial task framework
An industry-oriented framework describing 18 manipulation task categories and 55 evaluation cases, according to the publisher.
Bench2Dex research collaborators
Bench2Dex
A released Isaac Lab benchmark for simulated bimanual visuo-tactile manipulation. The authors report 12 hand embodiments, 26 tasks and about 1,300 teleoperated demonstrations.
NIST
NIST manipulation measurement research
Measurement research for grasping, manipulation and contact safety, including assembly task boards and repeatable evaluation methods.
YCB benchmarking project
YCB protocols and benchmarks
Object sets, task protocols and benchmark references for manipulation research.
RobotPerf contributors
RobotPerf
A robotics computing benchmark suite using ROS 2 as a common baseline.
OpenTeach / NYU research team
OpenTeach
Headset-based teleoperation and demonstration collection for several robot arms and dexterous hands.
OpenTeleVision research team
Open-TeleVision
Immersive robot teleoperation with active stereo visual feedback and hand-motion tracking.
BunnyVisionPro research team
BunnyVisionPro
Bimanual dexterous teleoperation with Vision Pro tracking and real-time retargeting.
HORA research team
HORA
In-hand object rotation using reinforcement learning and rapid motor adaptation.
DexMV research team
DexMV
Simulation and retargeting tools for learning dexterous manipulation from human videos.
DexPoint research team
DexPoint
Point-cloud reinforcement learning for generalizable dexterous manipulation.
PKU-EPIC
UniDexGrasp
Dexterous grasp proposal generation paired with goal-conditioned grasp policies.
DexDeform research team
DexDeform
Dexterous deformable-object manipulation with human demonstrations and differentiable physics.
DAPG research team
DAPG
Demonstration-augmented policy-gradient learning for dexterous hand manipulation.
Google Research / RoboPianist authors
RoboPianist
A MuJoCo benchmark for coordinated piano playing with two simulated robotic hands.
Farama Foundation
Gymnasium Robotics
MuJoCo robotics environments with Gymnasium interfaces, including dexterous hand tasks.
Isaac Lab project
Isaac Lab
GPU-oriented robot-learning framework with articulated-robot manipulation environments.
ManiSkill contributors
ManiSkill
SAPIEN-based manipulation simulation with GPU-parallel environments and robot-learning tools.
ARISE Initiative
robosuite
Modular MuJoCo simulation framework for robot manipulation and benchmark tasks.
ARISE Initiative
robomimic
A framework for training and evaluating manipulation policies from demonstrations.
Meta Research / DIGIT project
DIGIT Interface
Python interface for connecting to DIGIT vision-based tactile sensors and reading their output.
Tactile Gym research team
Tactile Gym 2.0
PyBullet learning environments using simulated tactile observations for contact tasks.
DexImit research team
DexImit
A pipeline for learning bimanual dexterous manipulation from monocular human videos.
DFKI Robotics Innovation Center
Hand Embodiment
A modular pipeline transferring captured human-hand motions to different robotic embodiments.
Pollen Robotics
AmazingHand
A printable four-finger research hand with mechanical files and control software.
Chestnut Robotics
Aero Hand Open
A tendon-driven research-hand design with assembly, firmware and simulation resources.
ORCA Hand
ORCA Hand Hardware
CAD, printable parts and assembly resources for ORCA hand variants.
Yale GRAB Lab
Yale OpenHand
CAD and build resources for a family of adaptive robotic hands and grippers.
LEAP Hand
LEAP Hand v2 Standard SDK
Python and ROS 2 interfaces with serial setup and calibration for LEAP Hand v2 Standard.
LEAP Hand
LEAP Hand in Isaac Lab
An Isaac Lab environment and sim-to-real examples for LEAP v1 single-axis in-hand reorientation.
ORCA Hand
ORCA Core
Python control package with calibration, tendon tensioning and joint-space commands.
ORCA Hand
ORCA Simulation
MuJoCo simulation environments for ORCA hands using the Gymnasium API.
ORCA Hand
ORCA Teleoperation
Capture-source and retargeting tools for operating ORCA hands from human hand motion.
PSYONIC
Ability Hand API
Manufacturer documentation, Python/C++ wrappers and simulation examples for the Ability Hand.
SCHUNK
SCHUNK SVH ROS Driver
SVH robot-hand driver with descriptions, trajectory control and simulation integration.
Shadow Robot
Shadow Hand ROS Interface
ROS interfaces and integration tools for the Shadow dexterous-hand family.
Shadow Robot
DEX-EE ROS Interfaces
ROS messages, services and robot descriptions for DEX-EE.
Wonik Robotics
Allegro Hand v5 ROS 2
Official ROS 2 control and feedback packages for the four-finger Allegro v5 hand.
DexRobot
DexCanvas
Dataset-loading tools for human hand-motion trajectories in Lance/Parquet formats.
DexRobot
DexRobot MuJoCo
DexHand simulation tools with collision-model options and tactile-sensor examples.
Sharpa Robotics
Sharpa Wave SDK
Libraries and API documentation for integrating Sharpa Wave into robotics applications.
Wuji Technology
Wuji SDK
Device discovery, streaming, recording and hand-retargeting interfaces for Wuji devices.
NYU Robot Learning
Inspire Hand Research SDK
Research-maintained Python tools for RH56 hand control and fingertip-target inverse kinematics.
PKU-EPIC / BODex research team
BODex
GPU-based dexterous grasp synthesis and approach-trajectory planning using bilevel optimization.
PKU-EPIC
DexGraspNet 2.0
Generative dexterous grasping tools for large synthetic cluttered scenes.
NVIDIA Research / University of Washington
TacSL
Visuotactile sensor simulation and policy-learning research for contact-intensive manipulation.
MandiZhao
DexMachina
Physics-based retargeting of human demonstrations into coordinated robot-hand manipulation of articulated objects.
dexwild
DexWild
Portable human demonstrations and robot data combined to train dexterous manipulation policies.
real-stanford
DexUMI
A wearable interface that records dexterous demonstrations and adapts their visual appearance for robot learning.
Meowuu7
DexTrack
Dexterous manipulation through reinforcement learning that tracks human-derived hand–object motion.
PKU-EPIC
UniDexGrasp++
A generalist-and-specialist training approach for dexterous grasping across object geometries.
tengyu-liu
GenDexGrasp
Contact-map-based grasp synthesis designed to generalize across robotic hand embodiments.
microsoft
UniGraspTransformer
Transformer policies distilled from specialist grasping policies to handle varied objects.
JYChen18
Dexonomy
Contact templates and a grasp taxonomy for synthesizing human-like dexterous grasp configurations.
facebookresearch
Sparsh
Self-supervised encoders for vision-based touch, with downstream evaluation through TacBench.
raunaqbhirangi
AnySkin
Python tools for collecting measurements from a magnetic tactile skin.
irmakguzey
T-Dex / Dexterity from Touch
Learning tactile representations from exploratory interaction and using them for dexterous manipulation.
ManipTrans
ManipTrans
Transfers dexterous bimanual human demonstrations to robots with residual learning.
Liang-ZX
DexHandDiff
Interaction-aware diffusion planning for dexterous manipulation.
NVlabs
DextrAH
Hand–arm grasping research that transfers privileged training information into visual policies.
Improbable-AI
Visual Dexterity
Visual learning for in-hand object reorientation with a D’Claw robot.
YuLiHN
DexDiffuser
A grasp sampler, evaluator and refinement pipeline conditioned on partial object point clouds.
MultiGrasp
MultiGrasp
Dexterous grasp generation and execution research for holding multiple objects in one hand.
4DVLab
DexH2R
A benchmark for dynamic human-to-robot handovers with dexterous hands.
PKU-MARL
Bi-DexHands
A collection of bimanual dexterous manipulation tasks for reinforcement-learning experiments.
dexsuite
Dex URDF
A collection of hand and object model descriptions for dexterous robotics workflows.
robot-descriptions
robot_descriptions.py
Python loaders for robot description packages, including models useful for hand experiments.
lixiny
manotorch
A differentiable MANO hand-model layer with anatomical hand utilities in PyTorch.
hassony2
manopth
A PyTorch implementation of the MANO hand-model layer for differentiable hand reconstruction.
geopavlakos
HaMeR
Transformer-based reconstruction of three-dimensional human hand meshes from images.
facebookresearch
Hand Tracking Toolkit
Tools for reading, visualizing and evaluating hand-tracking data and model outputs.
rolpotamias
WiLoR
Hand detection and three-dimensional hand reconstruction in varied visual scenes.
ThunderVVV
HaWoR
Reconstruction of world-space hand motion from egocentric video.
zc-alexfan
HOLD
Joint reconstruction of a hand and a held object from video.
ZhengdiYu
Dyn-HaMR
Reconstruction of interacting hand motion in four dimensions from moving-camera video.
ruka-hand
RUKA
A tendon-driven research hand with design, calibration and control resources.
ruka-hand-v2
RUKA-v2
A later RUKA design with wrist and finger-abduction capabilities and accompanying control tools.
SoutheastWind
Tactile SoftHand-A
Design resources for a tactile, underactuated hand with antagonistic tendon actuation.
midas-hand-org
MIDAS Hand API
Python control, homing and tactile-sensor tools for the MIDAS hand.
zhaohengyin
Lightning Grasp
A contact-field-based procedural grasp-generation engine used by DexterityGen.
facebookresearch
SPIDER
Physics-based retargeting of human motion into robot trajectories across supported embodiments.
malik-group
Do as I Do
A video-to-robot workflow combining reconstruction, retargeting and model-predictive control.
Rice-RobotPI-Lab
EgoInfinity
A reconstruction pipeline for recovering metric hand–object interaction from RGB video.
brave-eai
DexJoCo
A MuJoCo benchmark for coordinated dexterous manipulation and tool use.
srl-ethz
PoMDAR
A dexterous manipulation benchmark with simulation assets, printable objects and demonstration interfaces.
ycyao216
DexVerse
A research benchmark intended to cover multiple dexterous hand tasks and embodiments.
mimicrobotics
Smooth Operator / mimic retargeter
Retargeting research and tools for mapping human motion onto different dexterous hands.
iSEE-Laboratory
TypeTele
Manipulation-type-aware retargeting for dexterous teleoperation.
yunhaif
REGRIND
Retargeting-guided reinforcement learning for dexterous manipulation.
tylerlum
Get a Grip
A grasp dataset and learning pipeline connecting grasp proposals to hand–arm planning.
IRVLUTD
RobotFingerPrint
A common-coordinate representation for transferring grasp information across different robot hands.
google-deepmind
GeoMatch
Geometry-matching research for grasp synthesis across different robot embodiments.
4DVLab
DexGraspAnything
Physics-aware dexterous grasp generation across varied objects.
GeorgeWuzy
CEDex
Cross-embodiment dexterous grasping with a human-like contact representation.
hgaurav2k
HOP
Hand–object interaction pretraining followed by task-specific dexterous policy learning.
x-robotics-lab
DexScrew
Dexterous tool manipulation using a learned rotation prior and real-world sensing adaptation.
ZaynLeung
DexTele / Robotelop
Dual-arm retargeting and teleoperation code with physical-robot recordings and simulation examples.
WenjiaWang0312
EmbodMocap
Four-dimensional reconstruction of human motion and scene context from everyday recordings.
Tao-11-chen
ForeHOI
Feed-forward object reconstruction and pose estimation from partially observed hand–object interaction videos.
aidilayce
Follow My Hold
Hand–object reconstruction from video using geometric guidance.
jeffsonyu
ViTaM-D
Dynamic hand–object reconstruction using a distributed, force-aware contact representation.
BeingBeyond
TTP / H-Tac pretraining
Human-centric tactile pretraining and downstream dexterous manipulation learning.
ada-grasp-ctrl
CoorGrasp
Coordinated contact control for adapting a dexterous grasp under uncertainty.
ZhuoyangLiu2005
T-Rex
Tactile-reactive dexterous manipulation that uses touch to respond during task execution.
cypypccpy
ObjDex
Object-centric dexterous manipulation policies learned from human interaction motion.
Hly-123
FunGrasp
Functional grasping research for multiple dexterous hand designs.
GaTech-RL2
EgoVerse
Tools for processing, accessing and learning from egocentric human data for robot learning.
taeyunwoo
DexHOI2Sim
Physics-based evaluation of reconstructed or generated human hand–object interactions.
Farama-Foundation
D4RL / Adroit legacy benchmarks
Historical offline reinforcement-learning benchmarks, including Adroit dexterous-hand tasks.
haosulab
SAPIEN
A physics and rendering platform for articulated-object interaction and robotic manipulation.
nvidia-isaac
NVIDIA Video to Data
A pipeline connecting human video reconstruction with simulation assets and robot-training data.
open-dynamic-robot-initiative
TriFinger Simulation
Official simulation environments and models for the TriFinger robot family.
YuyangLee
TacThru
Combined tactile and visual perception for learning contact-rich manipulation.
armlabstanford
DenseTact reconstruction assets
Geometry and index files used to train shape reconstruction for the DenseTact optical tactile sensor.
facebookresearch
DIGIT sensor design
Manufacturing documentation and design files for the DIGIT vision-based tactile sensor.
AureleoPKU
DeCAL
A dexterous vision-language-action model combining contact-aware multimodal prediction and action generation.
LogosRoboticsGroup
TacPAC
Tactile prediction and execution-time action correction for contact-rich robot manipulation.
real-stanford
UMI-FT
A handheld interface and learning workflow for recording forceful manipulation without using a robot during capture.
gist-ailab
ManipForce
Force-guided policy learning with frequency-aware representations for contact-rich manipulation.
kushal2000
Play2Perfect
Dexterous play pretraining followed by task-specific fine-tuning for precise assembly.
tonyzhaozh
ACT / Action Chunking with Transformers
An imitation-learning implementation that predicts chunks of robot actions from demonstrations.
real-stanford
Diffusion Policy
A visuomotor policy-learning framework that generates action sequences through diffusion.
Robbyant
LingBot-VLA 2.0
A vision-language-action release with an action representation that includes dexterous hands and other robot subsystems.
physical-superintelligence-lab
HumDex
A humanoid dexterous manipulation stack combining teleoperation, data collection and policy learning.
SimarKareer
EgoMimic
Joint learning from egocentric human video and teleoperated robot data.
airoa-org
YUBI interface software
Operator-side software for a two-handed manipulation-data collection interface.
wangwen-banban
EMPIRE
Egocentric hand-motion forecasting with an explicit representation of the manipulation plan.
facebookresearch
GeoRT
Neural geometric retargeting for mapping human motion onto robotic hands.
neuroagents-lab
Tactile Genesis
A self-contained simulation snapshot for studying tactile sensors in dexterous task learning.
EDEM-AI
RoboSynChallenge
A dexterous manipulation benchmark connecting synthesized skills, training data and evaluation.
CEA-LIST
CoToGrasp
Contact-topology-conditioned grasp synthesis using a canonical workspace representation.
CEA-LIST
GOAG
A generative, object-agnostic planner for dexterous robot grasps.
TEA-Lab
DOGlove
A haptic force-feedback glove with hardware documentation, calibration and teleoperation scripts.
wengmister
BiDexHand
A biomimetic research hand with cable-and-pulley mechanisms, build instructions and control examples.
wayrise
DexGarmentLab
A dexterous garment-manipulation environment with demonstration generation and policy-learning resources.
open-dynamic-robot-initiative
TriFinger robot_fingers
Official drivers, examples and basic utilities for the TriFinger robot family.
NYU-robot-learning
DIME / Dexterous Imitation Made Easy
A demonstration-collection and imitation-learning system for dexterous hand–arm manipulation.
srl-ethz
Faive Gym
Isaac Gym environments and training tools for the Faive hand.
raunaqbhirangi
D’Manus interfacing library
Control and data-collection interfaces for the D’Manus hand with tactile sensing over the fingers and palm.
InternRobotics
TAG tactile-feedback glove
A glove combining motion capture and tactile feedback for dexterous teleoperation.
Taccel-Simulator
Taccel
A GPU simulator for vision-based tactile sensing and contact-rich robotics experiments.
midea-ai
DexVLA
A vision-language model with a diffusion-policy component for visuomotor robot learning.
facebookresearch
Digit Plexus
Hardware and software resources for combining tactile sensors and robot-hand control on a shared interface.
raunaqbhirangi
ReSkin sensor library
Python interfaces, firmware and supporting resources for collecting ReSkin magnetic tactile measurements.
junha-l
DextER
Language-conditioned dexterous grasp generation evaluated on instruction and grasp-taxonomy benchmarks.
facebookresearch
NeuralFeels
Object perception during in-hand manipulation using neural fields and combined vision and touch.
Farama-Foundation
Minari
Tools for storing, sharing and using offline reinforcement-learning datasets.