So100 Stack Green Cube: Robot Telemetry and Video Data

A dataset created using LeRobot, containing 15 episodes and 6,619 frames of telemetry and video data for a single task. It was authored by Beegbrain and last updated on June 6, 2025. The data is structured for training machine learning models in robotics.

Use Cases

• Training robot control policies based on telemetry and video observations.
• Developing computer vision models for robotic task understanding from video data.
• Benchmarking imitation learning or reinforcement learning algorithms on a physical stacking task.
• Analyzing robot kinematics and state transitions across 15 recorded episodes.

Strengths

• Contains 6,619 frames of data across 15 episodes, providing a sequence of observations.
• Includes video data recorded at 30 frames per second, suitable for temporal analysis.
• Data is structured into a single chunk of 1,000 episodes for consistent access.

Limitations

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

Provenance

Source

huggingface

Collection Method

Created using LeRobot.

Freshness

Last updated 2025-06-06 17:22:44; freshness should be verified.