st-tech/zozo-shift15m

SHIFT15M: Fashion-specific dataset for set-to-set matching with several distribution shifts

39
/ 100
Emerging

This project provides the SHIFT15M dataset, a specialized collection of fashion item images and related data for evaluating machine learning models. It helps data scientists and machine learning engineers working in e-commerce or fashion tech who need to test how well their recommendation or categorization systems perform when user preferences or item trends change over time. The dataset takes in fashion item images and associated metadata, and outputs labeled data splits tailored for various machine learning tasks like predicting item popularity or matching sets of outfits.

176 stars. No commits in the last 6 months.

Use this if you are developing machine learning models for fashion e-commerce and need a large, real-world dataset to test their robustness against shifts in data distribution, like changing fashion trends or user behavior.

Not ideal if you are looking for a dataset of general-purpose images or if your machine learning task is not related to fashion items or set-to-set matching.

fashion-e-commerce product-recommendation image-classification trend-analysis online-retail
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 13 / 25

How are scores calculated?

Stars

176

Forks

16

Language

Python

License

Last pushed

Oct 18, 2023

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/st-tech/zozo-shift15m"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.