IndoorCrowd: Multi-Scene Dataset for Human Detection and Tracking

IndoorCrowd is a multi-scene dataset designed for indoor human detection, instance segmentation, and multi-object tracking. It captures diverse challenges such as viewpoint variation, partial occlusion, and varying crowd density across four distinct campus locations. The dataset was created by author 'sebnae' and was last updated on 2026-04-02.

Use Cases

• Training human detection models based on diverse indoor viewpoints and crowd densities.
• Developing instance segmentation algorithms to handle partial occlusion in indoor scenes.
• Benchmarking multi-object tracking performance across four distinct campus locations.
• Researching privacy-preserving computer vision techniques using the blurred face data.

Strengths

• Captures diverse challenges including viewpoint variation, partial occlusion, and varying crowd density.
• Explicitly blurs faces to preserve privacy, making it suitable for safe research.
• Covers four distinct campus locations (ACS-EC, ACS-EG, IE-Central, R-Central).

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and exact size are unknown, which may limit suitability assessment.

Provenance

Source

huggingface

Collection Method

Captured across four distinct campus locations (ACS-EC, ACS-EG, IE-Central, R-Central).

Time Range

null

Freshness

Last updated 2026-04-02 12:11:52.

Geography

Campus locations (ACS-EC, ACS-EG, IE-Central, R-Central).