MotIF-1K: 1,022 Robotic Motion Demonstrations Across 13 Tasks

MotIF-1K is a robotics motion dataset containing 1,022 demonstrations across 13 task categories. It is used to benchmark and fine-tune vision-language models for motion-based success detection. The dataset was created by Voxel51 and was last updated on the Hugging Face platform in May 2026.

Use Cases

• Fine-tuning vision-language models for motion understanding based on video demonstrations.
• Benchmarking model performance on motion-based success detection across 13 task categories.
• Analyzing the relationship between task instructions, motion descriptions, and visual trajectory visualizations.

Strengths

• Contains 1,022 demonstrations, providing a substantial number of examples for model training.
• Covers 13 distinct task categories, offering diversity in motion types.
• Each demonstration includes multiple data modalities: video, trajectory visualizations, and text instructions.

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

Voxel51 via Hugging Face.

Collection Method

Likely contains recorded robotic demonstrations and associated annotations.

Freshness

Last updated 2026-05-26 21:06:48; freshness should be verified.