Mouse Behavior Videos and Pose Data for Acute and Chronic Stress Models

411 open field recordings of mice include raw top-view .mp4 video files and corresponding .csv pose estimation data generated with DeepLabCut. The data originates from multiple experiments conducted by Lukas von Ziegler at ETH Zurich. A metadata file links the pose estimation files, located in subfolders, to their corresponding video files.

Use Cases

• Training pose estimation models based on top-view video recordings of mice.
• Analyzing behavioral patterns linked to acute and chronic stress models mentioned in the description.
• Benchmarking multi-animal tracking algorithms using raw video and corresponding pose data.
• Studying locomotion and posture in an open field environment across different experimental groups.

Strengths

• 411 distinct recordings provide a substantial number of behavioral samples.
• Includes both raw video files and pre-processed pose estimation data, enabling method validation.
• Metadata file explicitly links video and pose data across multiple experiment subfolders.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for pose estimation data is unknown, which may limit suitability assessment.
• Last update date is unknown; freshness unverified.

Provenance

Source

ETH Zurich (Lukas von Ziegler)

Collection Method

Recorded from multiple experiments; pose estimation performed with DeepLabCut.