Clothing Dataset Full
Dataset of clothing images including more than 11,500 images divided into 20 different categories, ideal for training visual classification models in fashion and e-commerce.
Approximately 11,500 images in various formats (JPEG, PNG), divided into 20 classes
CC0: Public Domain
Description
The Clothing Dataset Full contains over 11,500 images of clothing items organized into 20 classes. This dataset is ideal for training computer vision models specialized in clothing recognition and classification.
What is this dataset for?
- Develop visual recognition systems for clothing items
- Improving product recommendations in e-commerce
- Create image classification tutorials and courses
Can it be enriched or improved?
Yes, it is possible to add additional annotations such as detailed descriptions, style tags, or contextual metadata. The extension to segmentation or the recognition of images in context is also possible.
🔎 In summary
🧠 Recommended for
- AI developers
- Vision researchers
- Data science students
🔧 Compatible tools
- TensorFlow
- PyTorch
- FastAI
- OpenCV
💡 Tip
Use class diversity to train models that are robust to different styles and textures.
Frequently Asked Questions
Does this dataset contain annotations for each image?
Yes, each image is associated with a class corresponding to a type of clothing item.
What is the average resolution of the images?
The images are varied in resolution, often adapted to machine learning use, around 224x224 pixels or more.
Is this dataset suitable for recognition in real conditions (e.g. photos taken in store)?
Primarily, these are isolated images, but can be completed for in-context recognition with additional annotations.
