/Fashion_Recommendation_System

This project is about recommending similar fashion products when we do an image search using some of the products we want to buy.

Primary LanguageJupyter Notebook

Fashion Recommendation System

This is a Fashion Recommendation System that leverages image similarity search techniques to provide apparel recommendations based on visual preferences. The system applies a preprocessing pipeline, including resizing input images to 299x299 dimensions and extracting 2048-dimensional embeddings using a pre-trained Xception network.

Features

  • Image Similarity Search: The system utilizes image embeddings and cosine similarity to perform efficient similarity matching between fashion apparel items.
  • Dimensionality Reduction: Principal Component Analysis (PCA) is applied to reduce the dimensionality of the embeddings to 128, facilitating faster similarity computations.
  • Innovative Recommendation Algorithm: The system implements a novel recommendation algorithm that suggests the most similar fashion apparel items based on image similarity.
  • User-Friendly Interface: The Fashion Recommendation System is deployed using the Streamlit app, providing users with an intuitive and interactive interface to explore and receive personalized fashion recommendations.

Requirements

To run the Fashion Recommendation System, ensure that you have the following dependencies installed:

  • Python (version 3.7 or higher)
  • Streamlit (version 1.2.0)
  • TensorFlow (version 2.7.0)
  • Keras (version 2.6.0)
  • NumPy (version 1.19.5)
  • OpenCV (version 4.5.4.60)
  • scikit-learn (version 1.0.1)
  • pandas (version 1.3.4)
  • tqdm (version 4.62.3)
  • plotly (version 5.4.0)
  • matplotlib (version 3.5.1)

Usage

  1. Clone the repository: git clone https://github.com/your-repo.git
  2. Install the required dependencies: pip install -r requirements.txt

Examples

To provide a visual representation of the Fashion Recommendation System, below are some screenshots of the Streamlit app:

[Insert screenshots of the app interface here]

License

This project is licensed under the MIT License.

Acknowledgments

  • The Xception network is utilized as a pre-trained model to extract image embeddings.
  • The Streamlit framework is used for deploying the Fashion Recommendation System to production.

Contact

For questions or inquiries, please contact [srv.ale52@gmail.com].