Drosophila Object Manipulation Tracking Data and Videos

Raw and processed behavioral tracking data from wild-type Drosophila experiments by Durrieu et al. 2026. The dataset includes per-animal HDF5 tracking files, processed behavioral metrics, and grid videos of experiments. It was published on May 12, 2026, focusing on object manipulation and affordance learning.

Use Cases

• Train pose estimation models based on full-body keypoint pose files.
• Analyze object interaction dynamics based on ball-pushing metrics.
• Validate behavioral tracking algorithms using raw centroid tracking data.
• Generate synthetic training data for animal behavior models using provided grid videos.

Strengths

• Includes multiple data modalities: raw video recordings, processed coordinates, and behavioral metrics.
• Provides data in both raw HDF5 format and processed Feather format for different analysis stages.
• Dataset is associated with a specific 2026 research publication, providing context.

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

Durrieu, Matthias

Collection Method

Experimental recordings and processing from Drosophila behavior studies.

Freshness

Last updated 2026-05-12 14:44:23; freshness should be verified.