Image-to-Image-Search: A Python repository from sethuiyer

SmartSearch is a reverse image search engine which finds similar images by generating captions and comparing those captions. Powered by Tensorflow and ElasticSearch.

Update: Mirror of this repository in codeberg: Image-to-Image-Search

Update: REST APIs for image caption generator and image search are added! ( Shoutout to xc0d3rz )

Demo

Deep Reverse Image Search Engine - YouTube demos the core functionality.

Branches

Docker Image

Basic Implementation

Packages Required:

Anaconda
Keras with Tensorflow Backend (Python 3.6)
Elastic Search and elasticsearch-py (Elastic Search 6.0)

For more, check out requirements.txt

Pre-trained models

Flickr-8k LSTM weights (flickr8k_cnn_lstm_v1.p)

Download this and paste it inside models folder.

Output

Tips

Install elasticsearch and always check if elastic search process is running before launching server.py or index_database.py.
Instead of using the upload functionality, paste all your images inside static/img folder followed by python index_database.py to index all those images.
If you want to delete the indexed images, do sh delete_index.sh

Thanks

Thanks to Materialize CSS for the aesthetics, Pretty Printed for the flask tutorials and Andrej Karpathy for wonderful captioning library. Shoutout to xc0d3rz for the REST API inclusion!

sethuiyer/Image-to-Image-Search