Exylos Bimanual Robot Spill Cleanup with Full Modalities Sample

A sample dataset from ExylosAi provides synchronized data for bimanual robot manipulation on a tabletop spill cleanup task. It includes dual-arm state trajectories, seven RGB camera streams, depth maps, segmentation masks, object poses, phase annotations, and a cleanup success metric. The dataset was last updated on May 17, 2026.

Use Cases

• Training imitation learning models based on synchronized dual-arm Panda state and action trajectories.
• Developing computer vision models for scene understanding using the seven RGB camera streams and per-frame depth maps.
• Evaluating segmentation algorithms on per-frame segmentation masks for objects and spills.
• Benchmarking robotic task performance using the objective cleanup success metric based on remaining spill fraction.
• Analyzing task phases and temporal structure in manipulation using the provided phase annotations.

Strengths

• Includes multiple synchronized data modalities: RGB video, depth, segmentation, poses, and robot state.
• Provides an objective performance metric (cleanup success based on remaining spill fraction).
• Contains phase annotations for temporal task analysis.

Limitations

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

Provenance

Source

ExylosAi via Hugging Face

Freshness

Last updated 2026-05-17 17:13:15; freshness should be verified.