hanxiao / HM-Fashion-image-neural-search

H&M Fashion Image similarity search with Weaviate and DocArray

This example shows how to do image similarity search using DocArray and Weaviate as Document Store.

Download data

You need to download the fashion image data from the H&M dataset on Kaggle. You can download it and put in the right folder using:

$ mkdir data
& cd data
$ kaggle competitions download -c h-and-m-personalized-fashion-recommendations
$ unzip h-and-m-personalized-fashion-recommendations.zip

Start up Weaviate

You need to start up Weaviate using docker compose up. Weaviate will be running on http://localhost:8080

Embed, store and query

You can run the Jupyter notebook to embed, store and query fashion image data using ResNet50, DocArray and Weaviate.