Skip to content

Latest commit

 

History

History
77 lines (66 loc) · 3.2 KB

README.md

File metadata and controls

77 lines (66 loc) · 3.2 KB

Simple Image Search

The Image Search App is a simple tool that allows users to search for specific images or find visually similar images within a database. It utilizes advanced computer vision techniques to extract features from images and perform similarity matching, enabling various applications such as searching for faces or finding similar products using images.

Features

  • Image Search: Users can search for specific images by uploading an image or providing a query description. The app matches the query against the image database and retrieves relevant results based on image similarity.
  • Similar Image Retrieval: Users can find visually similar images to a given query image within the database. The app employs advanced image embedding techniques and similarity algorithms to identify images with comparable visual characteristics.

Installation

Image Search depends on PyTorch, Faiss and Streamlit. Below are quick steps for installation.

conda create --name img_search python=3.8.10
conda activate img_search
conda install pytorch torchvision pytorch-cuda=11.8 -c pytorch -c nvidia
pip install ftfy regex tqdm
pip install git+https://github.com/openai/CLIP.git
git clone https://github.com/tanluuuuuuu/Simple-Image-Search.git
cd Simple-Image-Search
pip install -r requirements.txt

Usage

  1. Activate the environment:
conda activate img_search
  1. Navigate to the project directory:
cd Simple-Image-Search
  1. Start the application:
streamlit run scr/main.py
  1. It will automatically open your web browser. If it doesn't, open your web browser and go to http://localhost:8501 or http://0.0.0.0:8501
  2. Upload an image or provide a query description to search for similar images.
  3. View the search results and explore visually similar images.

Technologies Used

  • Programming Language: Python
  • Computer Vision Libraries: Pillow, PyTorch
  • Deployment: Streamlit

Docker

You can also run the Image Search App using Docker. Follow the steps below:

  1. Install Docker on your machine.
  2. Clone the repository:
git clone https://github.com/tanluuuuuuu/Simple-Image-Search.git
  1. Navigate to the project directory:
cd Simple-Image-Search
  1. Build the Docker image:
docker build -t img_search .
  1. Run the Docker container:
docker run -p 8501:8501 img_search
  1. Open your web browser and go to http://localhost:8051 or http://0.0.0.0:8501
  2. Provide database path then upload an image to search for similar images.
  3. View the search results and explore visually similar images.

Contributing

Contributions to the Image Search App are welcome! If you find any bugs or have suggestions for new features, please open an issue or submit a pull request. Follow the guidelines in the CONTRIBUTING.md file for contributing to the project.

Contact

For any inquiries or feedback, please contact [[email protected]]

Acknowledgments

I would like to thank the contributors and the open-source community for their support and contributions to the Image Search App.

Note: The Image Search App is for demonstration purposes only and may require additional configuration or customization to suit specific use cases or production environments.