Underwear Occlusion Image Dataset
V1.0Latest Update: | 2025-10-15 | Samples: | 20000 records |
Samples: | 20000 records | ||
File Size: | 1.8G | Format: | JPG/PNG/JSON |
Format: | JPG/PNG/JSON | ||
Data Domain: | Image | Holder: |  MOBIUSI INC |
Holder: | MOBIUSI INC | ||
Industry Scope: | Product Display | Visual Recognition | E-commerce Analysis | ||
Applications: | Image Classification | Object Detection | ||
Brief Introduction
The retail e-commerce industry is currently facing challenges with visual recognition systems that struggle to accurately identify and classify products that are partially occluded. Existing solutions often overlook the nuances of occluded images, leading to poor classification results. This dataset aims to address these shortcomings by providing a comprehensive collection of images featuring occluded underwear, thus enabling more robust training of models for visual recognition tasks. The data is collected using high-resolution cameras in controlled lighting conditions to ensure clarity and detail. Quality control measures include multi-round annotations, consistency checks among annotators, and expert reviews to maintain a high standard. The images are organized in a structured directory format and stored in JPG format for easy access and processing.
Sample Examples
| Image | File Name | Resolution | Target Quantity | Target Type | Target Visibility | Lighting Conditions | Background Clarity | Annotation Quality |
|---|---|---|---|---|---|---|---|---|
 | b8f3a32d317e7f598d26dc93414e67f4.png | 1704*1500 | 1 | bra | fully visible | bright | blurry | 4 |
 | 2b0e68789be9095f4ce6a743f2681721.png | 1021*1500 | 1 | bra | fully visible | bright | clear | 5 |
 | 74956ad949d9d935f8a0d83359fddc7d.png | 867*1500 | 1 | bra | fully visible | bright | clear | 5 |
 | 948d24308fabe9e288ae6f1ec7d0b6a3.png | 1517*1500 | 1 | bra | fully visible | bright | blurry | 5 |
 | 517659d80f2a8e47b67eeda7e612b9b6.png | 1440*1500 | 1 | bra | fully visible | bright | clear | 5 |
Data Structure
| Field | Type | Description |
|---|---|---|
| file_name | string | File name |
| quality | string | Resolution |
| object_count | int | The total number of targets identified in the image. |
| object_type | string | The type of each target in the image, such as bra, panties, etc. |
| object_visibility | string | The degree of target visibility in the image, such as fully visible, partially occluded, etc. |
| lighting_conditions | string | The lighting conditions in the image, such as bright, dark, artificial light. |
| background_clarity | string | The clarity of the image background, such as clear or blurry. |
| annotation_quality | int | The accuracy of the image annotations, rated on a scale of 1 to 5. |
Compliance Statement
| Item | Content |
|---|---|
| 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 |
Can't find the data you need?
Post a request and let data providers reach out to you.