Commercial Retail Theft Action Recognition Dataset

#action recognition #behavior analysis #supermarket #convenience store #clothing store

500 records
1.6G
MP4/JSON
CC-BY-NC-SA 4.0
MOBIUSI INC

Updated:2026-02-04

AI Analysis & Value Prop

With the rapid development of the commercial retail industry, theft has caused significant losses to merchants. Existing monitoring systems often find it difficult to detect complex theft behavior in real-time, resulting in merchants being unable to take timely measures after losses occur. This dataset aims to assist researchers and developers in building efficient action detection models to meet this challenge by providing a large number of real commercial retail theft behavior videos. Data collection is conducted by deploying surveillance cameras in supermarkets, convenience stores, and clothing stores, capturing the behavior of real customers and staff. To ensure data quality, multi-round annotations and consistency checks are used, combined with expert reviews to ensure the accuracy of the annotations. The data is stored in MP4 format, and video files are organized by date and location. The dataset contains 5000 video samples with a total file size of about 10GB.

Dataset Insights

Sample Examples

a2d636c9**.mp4|1280*720|7.66 MB

Technical Specifications

Field	Type	Description
file_name	string	File name
duration	string	Duration
quality	string	Resolution

Compliance Statement

Authorization Type	CC-BY-NC-SA 4.0 (Attribution–NonCommercial–ShareAlike)
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

Frequently Asked Questions

What is the Retail Theft Action Recognition Dataset?: The Retail Theft Action Recognition Dataset is a video dataset aimed at recognizing theft behavior in retail environments, specifically for applications in the security industry.

In which industry is this dataset primarily used?: This dataset is primarily used in the security industry to enhance the recognition and monitoring capabilities of theft behavior in retail environments.

What modalities of data does this dataset contain?: The Retail Theft Action Recognition Dataset contains video modalities of data, focusing on video understanding and analysis.

How can this dataset be used in the security field?: This dataset can be used to train and test automated video surveillance systems to recognize and detect suspicious theft behaviors in retail stores, thereby enhancing the efficiency of security measures.

Why do retail environments need this type of dataset?: Retail environments are complex and densely populated, prone to theft incidents. Using such datasets can help improve the speed and accuracy of recognizing and responding to theft, thereby reducing losses.

Can't find the data you need?

Post a request and let data providers reach out to you.

Get this Dataset

Verified for Enterprise Use

Cite this Work

@dataset{Mobiusi2025,
  title={Commercial Retail Theft Action Recognition Dataset},
  author={MOBIUSI INC},
  year={2025},
  url={https://www.mobiusi.com/datasets/e55e5463a5338daad0df2005bd11f9cb?dataset_scene_id=10},
  urldate={2025-09-15},
  keywords={commercial retail, theft action recognition, security dataset, action detection, video surveillance},
  version={1.0}
}

Using this in research? Please cite us.

placeholder

Scene

Domain

Modality

Task