Eval2 Bgr G: Robot Demonstration Episodes

4 episodes of robot demonstration data created using LeRobot. The dataset contains 2400 frames recorded at 30 frames per second and is split into a single training set. It was uploaded by robot-learning-group47 and last updated on May 14, 2026.

Use Cases

• Training robot imitation learning models based on demonstration episodes.
• Benchmarking robot policy performance based on recorded video and state data.
• Developing computer vision models for robot state estimation based on visual frames.
• Analyzing robot task execution patterns based on sequential episode data.

Strengths

• Contains 2400 frames of recorded robot data.
• Includes 4 complete demonstration episodes.
• Recorded at a standard 30 frames per second for video consistency.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

robot-learning-group47 via Hugging Face

Collection Method

Created using LeRobot.

Freshness

Last updated 2026-05-14 13:38:35; freshness should be verified.