Sanket - ISL Translator: Continuous Sign Language Recognition

This repository contains the codebase for the Sanket - ISL Translator, a final year project focused on continuous sign language recognition using the INCLUDE dataset.

Project Structure

• Sanket - ISL Translator/: The main inference application and model architecture for running the demo.
• Sanket Training Scripts/: Scripts and instructions to train or fine-tune the model from scratch.
• Dataset/: Instructions and scripts to download and prepare the INCLUDE dataset.

Setup & Running the Demo

1. Clone this repository:

   git clone https://github.com/sanket/Sanket-ISL-Translator.git
   cd Sanket-ISL-Translator

2. Create and activate a virtual environment: Windows:

   python -m venv venv
   venv\Scripts\activate

   Linux/Mac:

   python3 -m venv venv
   source venv/bin/activate

3. Install dependencies:

   pip install -r "Sanket - ISL Translator/requirements.txt"

4. Download the pre-trained weights from Hugging Face (see section below).
5. Run the demo:

   cd "Sanket - ISL Translator"
   python app.py

📥 Downloading Model Weights

The trained model weights are hosted on Hugging Face.

1. Download model_best.pth from the Hugging Face Repository.
2. Place the downloaded model_best.pth file inside the Sanket - ISL Translator/weights/ folder.