Sign Language Prediction

A deep learning-based system for real-time American Sign Language (ASL) alphabet recognition using a webcam and Convolutional Neural Networks (CNN).

🚀 Overview

This project enables live ASL hand gesture recognition, translating static signs into text using a custom-trained CNN model. It includes tools for data collection, preprocessing, training, and real-time inference.

📁 Repository Structure

• collect_imgs.py – Capture labeled gesture images via webcam.
• create_dataset.py – Process and convert images into a structured dataset.
• train_classifier.py – Train the CNN classifier.
• inference_classifier.py – Perform real-time ASL prediction.
• graph.py – Visualize training metrics (accuracy/loss).
• model.p – Saved trained model.
• data.pickle – Serialized labels and data features.
• LICENSE – Project license information.

📘 Documentation

Here is the report by me ----> Project Report

📄 License

This project is licensed under the MIT License 📝

Questions and doubts are welcomed here.