This is a simple reverse image search tool that uses the Streamlit framework. It allows users to find similar images from a specified directory.
This tool can help identify similar images. This tools uses EfficientNet-B0 for vector embedding and Chromadb(vector database) for vector storage and retrieval. EfficientNet-B0 is a lightweight convolutional neural network architecture that is designed to be efficient and accurate for image recognition tasks.
To use the reverse image search engine, simply follow these steps:
# clone the repo
git clone https://github.com/tikendraw/reverse-image-search.git
# go inside
cd reverse-image-search
# install with pip
pip install .
bash command
# bash command
image_search
or just run the launch.py
file
python launch.py
The reverse image search engine is built using the following technologies:
Future updates:
If you would like to contribute to the reverse image search engine, please feel free to open a pull request.
@misc{tan2020efficientnetrethinkingmodelscaling,
title={EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks},
author={Mingxing Tan and Quoc V. Le},
year={2020},
eprint={1905.11946},
archivePrefix={arXiv},
primaryClass={cs.LG},
url={https://arxiv.org/abs/1905.11946},
}
The reverse image search engine is licensed under the MIT License.
I hope this is helpful!