MOBIUSI INCThis dataset contains images of people fighting and walking in public scenes, together with corresponding annotations, for distinguishing these two representative behaviors. Each image record includes a unique identifier, image and annotation file paths, a category ID, and image dimensions and format to support data organization and model training. It is suitable for public-area incident monitoring, video analytics development, and evaluation of visual recognition models.
| Field | Type | Description |
|---|---|---|
| file_name | string | File name |
| quality | string | Resolution |
| image_id | string | A unique identifier for the image record. |
| image_path | string | The file path to the image. |
| label_path | string | The file path to the corresponding annotation. |
| category_id | array | The identifier for the image's behavior category. |
| image_width | number | The width of the image in pixels. |
| image_format | string | The file format of the image. |
| image_height | number | The height of the image in pixels. |
| Authorization Type | Proprietary - Commercial AI Training License (No Redistribution) |
| Commercial Use | Requires exclusive subscription or authorization contract (monthly or per-invocation charging) |
| Privacy and Anonymization | No PII, no real company names, simulated scenarios follow industry standards |
| Compliance System | Compliant with China's Data Security Law / EU GDPR / supports enterprise data access logs |

Post a request and let data providers reach out to you.
@dataset{Mobiusi2026,
title={Public-Scene Fighting and Walking Behavior Image Recognition Dataset},
author={MOBIUSI INC},
year={2026},
url={https://www.mobiusi.com/datasets/d8ac55d700ac8be27d8b060854cd1ec2},
urldate={2026-11-04},
keywords={human behavior recognition dataset,fighting behavior images,walking behavior recognition,public scene vision data,abnormal behavior detection,security image annotations,behavior classification training data,people monitoring},
version={1.0}
}Using this in research? Please cite us.