Mobiusi/Down-Jacket-Vest-Occlusion-Image-Dataset

Down Jacket Vest Occlusion Image Dataset

In the current retail e-commerce industry, the challenge of accurately recognizing clothing items that are partially occluded (e.g., due to overlapping clothing or missing parts) is significant. Existing datasets often do not cover such scenarios adequately, leading to insufficient training for models. This dataset aims to address the technical issue of recognizing and segmenting images of down jacket vests that are obscured by other garments, fulfilling a crucial business need for e-commerce platforms to improve visual search and recommendation systems. The dataset was collected using high-quality cameras in various retail environments, ensuring diverse lighting and angles. To maintain high quality, we implemented multiple rounds of annotation with consistency checks and expert reviews. The images are stored in JPG format, organized by categories based on the occlusion types, facilitating easy access and use.

Technical Specifications

Field	Type	Description
file_name	string	File name
quality	string	Resolution
occlusion_level	string	The degree of occlusion of the down vest in the image, such as: no occlusion, partial occlusion, severe occlusion.
background_complexity	string	The complexity of the image background, such as: simple, average, complex.
texture_pattern	string	The main material texture of the down vest, such as: smooth, rough, fluffy.

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

Source & Contact

If you need more dataset details, please visit Mobiusi. or contact us via contact@mobiusi.com