Dataset · Slep
Clothing images grouped by product
Slep/LAION-RVS-Fashion
Millions of product photos from LAION with text descriptions and tabular data for visual clothing search.
What's inside
Inside are 1 to 10 million records combining images, English texts, and tabular metadata. Images are selected from LAION corpora (2B EN, 2B MULTI TRANSLATED, and 1B NOLANG TRANSLATED) and grouped by extracted product IDs. The dataset was prepared by CRITEO AI Lab and ENPC researchers for visual search tasks with clarifying instructions.
What you can do with it
- Training visual clothing search systems by photo and text refinements
- Training multimodal models for matching product cards
- Evaluating e-commerce search algorithms on the open LRVS-F benchmark
- Languages
- English
- Size
- 1M–10M rows
- Downloads this month
- 281
- Likes
- 31
Models trained on it
How to load it
from datasets import load_dataset
ds = load_dataset("Slep/LAION-RVS-Fashion")