Fractal Nils: Robot Policy Learning Data via Zero-Shot Labeling

A dataset associated with the 2024 paper 'Scaling Robot Policy Learning via Zero-Shot Labeling with Foundation Models' presented at the 8th Annual Conference on Robot Learning. The dataset was uploaded by author 'holgerson' to Hugging Face on October 28, 2024. Its specific content, size, and structure are not detailed in the provided metadata.

Use Cases

• Training robot control policies based on data labeled using foundation models as described in the paper.
• Benchmarking zero-shot labeling techniques for robotic manipulation tasks.
• Studying the scalability of robot learning methods using pre-trained vision-language models.

Strengths

• Associated with a peer-reviewed conference paper from CoRL 2024.
• Uploaded to a major data platform (Hugging Face) in late 2024.

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, file formats, and license are unknown, which may limit suitability assessment.

Provenance

Source

Paper authors: Nils Blank, Moritz Reuss, Marcel Rühle, Ömer Erdinç Yağmurlu, Fabian Wenzel, Oier Mees, Rudolf Lioutikov.

Collection Method

Likely generated or labeled for robot policy learning experiments using foundation models, as per the associated paper.

Freshness

Last updated 2024-10-28 08:27:32; freshness should be verified.