Skip to content

Yttrium-32/feature-find

Repository files navigation

Feature Find

A web application written in Django that leverages the power of ResNet to reverse image search through your local directory of images.

Setup

This project recommends settings up with uv but pip may also be used.

Before the application can search through images, it needs to index them. If you're doing this for the first time with a large of images, this may take a while.

To begin, create a directory named photos/ and add your images to it.

Then, run the indexer:

uv run -m modules.index --dataset ./photos/

This will create a file called index.hdf5 at the root on the project. You are now ready to reverse image search!

If you wish to develop this application please install the pre-commit hook. You haven't launched the application even once, please manually add the pre-commit by running uv add pre-commit.

Install pre-commit hook

uv run pre-commit install

Usage

Start Django server:

uv run manage.py runserver

This will start the web app at 127.0.0.1:8000 by default.

About

Local reverse image search using neural networks

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published