humanoidsdata.com

Search

Search datasets, articles, and glossary terms for humanoids and embodied AI.

Training data for humanoid robots and embodied AI

Browse humanoid robot and embodied AI training datasets. Explore source-backed articles and glossary definitions covering robot learning, hardware, control, and simulation.

Get new datasets, articles, and glossary entries in your inbox.

Latest datasets

View all →
A mosaic of AIRoA MoMa 5k Human Support Robots performing mobile manipulation tasks

AIRoA MoMa 5k

AIRoA MoMa 5k contains 1,184,259 successful primitive-action episodes spanning 5,025.1 hours and 180,905,084 frames of teleoperated mobile manipulation with 44 Toyota Human Support Robots across five sites. The public, train-only LeRobot v3.0 release covers 68 main short-horizon task templates and combines head- and hand-camera RGB video with robot state and actions, end-effector pose, wrist force-torque history, low-level servo telemetry, and hierarchical task metadata.

View dataset →
A mosaic of RealOmni household manipulation demonstrations

10Kh RealOmni-Open Dataset

The 10Kh RealOmni-Open Dataset contains more than 13,000 hours and 5 million clips of bimanual human demonstrations captured across over 10,000 real household scenarios. Collected from more than 3,000 contributors with GenDAS grippers, it covers 30 manipulation skills in 10 scenario groups, including clothing, clutter organization, kitchen cleaning, and shoe handling. The 95 TB MCAP release pairs 1600 × 1296 fisheye video at 30 fps with reconstructed end-effector trajectories, gripper state, 6-axis IMU readings, and tactile-array signals.

View dataset →
A bimanual dexterous robot folding a paper airplane

Robotic Origami Challenge

The Robotic Origami Challenge dataset contains 682 real-world teleoperated demonstrations of a bimanual dexterous robot folding a traditional six-fold paper airplane. Its 51 collection seasons comprise more than 4.7 million frames at 30 fps, with 65-dimensional robot states and actions, joint torques, 10-fingertip force-torque signals, and six synchronized head, wrist, and tactile video streams. The release uses LeRobot v3.0, with many seasons also available in v2.1, for imitation learning, visual-tactile representation learning, and long-horizon deformable-object manipulation.

View dataset →
A mosaic of synchronized HiFi-UMI manipulation demonstrations

HiFi-UMI-2K

HiFi-UMI-2K is a 2,000-hour release of robot-free, bimanual manipulation demonstrations collected with the portable HiFi-UMI system. Each episode includes six microsecond-synchronized camera views, calibrated bimanual trajectories, gripper states, language annotations, subtask boundaries, and per-sample quality metadata. The data is automatically reconstructed and validated through simulation replay, then exported in a LeRobot v3-style Parquet and MP4 format for manipulation-policy pre-training and post-training.

View dataset →

Latest articles

View all →