Project Title: Foodie lens

Foodie Lens is a machine learning-based application designed to identify various traditional Yoruba dishes from images and provide the corresponding recipes and cooking instructions.

Try it out here: https://huggingface.co/spaces/Aramide/foodie_lens

Getting Started

These instructions will help you get a copy of the project up and running on your local machine

Installation

• Clone the repository git clone https://github.com/Adebesin-Aramide/Foodie_Lens.git cd foodie-lens-huggingface

• Install dependencies pip install -r requirements.txt

• Run the application locally streamlit run app.py

Deployment on Hugging Face

The application is deployed on Hugging Face Spaces. To deploy it:

1. Create a Hugging Face account if you don't have one.
2. Create a new Space and select Streamlit as the SDK.
3. Push your code to the new Space's repository.
4. Hugging Face will automatically build and deploy the application.

Usage

• Upload an image: Use the file uploader to select an image of a traditional Yoruba dish (limited to five dishes for now: asaro, ewagoyin, ekuru, moimoi, eforiro).
• View the prediction: The application will classify the dish and display the predicted class.
• Get the recipe: The application will provide the recipe and cooking instructions for the identified dish.

Data Collection

The dataset for this project was collected by web scraping images of traditional Yoruba dishes from Google. The images were then labeled and used to train the deep learning model.

Model

The custom model is based on a simplified version of VGG (TinyVGG) built using PyTorch. The model architecture includes two convolutional blocks followed by a classifier.

Tools and Technologies Used

• Python: The core programming language used for the project.
• PyTorch: A deep learning framework used to build and train the custom TinyVGG model.
• Streamlit: A web application framework used to create the interactive interface for the project.
• Docker: A containerization platform used to package the application for deployment.
• Hugging Face: A platform used to deploy the Streamlit application.
• Selenium: A Python library used for web scraping to collect image data.

License

This project is licensed under the MIT License.

Contact

For any inquiries or questions, please contact me via email: [email protected]

Note: This project is continuously being updated to include more classes of food. Stay tuned for more updates!