HORA: Hand–Object to Robot Action Dataset for Cross-Embodiment Learning

HORA is a large-scale multimodal dataset that converts human hand–object interaction demonstrations into robot-usable supervision. It combines HOI-style annotations like MANO hand parameters and object pose with embodied-robot learning signals such as end-effector trajectories under a unified canonical action space. The dataset was created by HORA-DB and last updated on Hugging Face in March 2026.

Use Cases

• Train robot manipulation policies based on human hand–object interaction demonstrations.
• Develop cross-embodiment learning models using unified canonical action space annotations.
• Analyze contact dynamics between hands and objects using provided contact annotations.
• Generate synthetic robot trajectories from human demonstrations for simulation.
• Benchmark models for translating visual observations to robot end-effector control.

Strengths

• Large-scale dataset designed for cross-embodiment learning.
• Multimodal annotations include MANO hand parameters, object pose, and contact data.
• Provides embodied-robot learning signals like robot observations and end-effector trajectories.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and exact size are unknown, which may limit suitability assessment.
• Data may reflect bias inherent to the specific collection methodology and source platform.

Provenance

Source

HORA-DB on Hugging Face.

Collection Method

Likely involves recording and annotating human hand–object interactions for robot learning.

Time Range

null

Freshness

Last updated 2026-03-16 10:40:50.

Geography

null