Skip to content

Speccco/Real-Time-Sign-Language-Classifier

Repository files navigation

📷 Real-Time Sign Language Classifier This app uses MediaPipe to detect hand landmarks and a trained neural network (best_model.keras) to classify American Sign Language (ASL) gestures in real time.

How It Works 🖐️ Detects your hand using the webcam 📌 Extracts hand landmarks (x, y, z points) 🧠 Passes them to a trained model 🗣️ Predicts the sign label (e.g., "Hello", "Thanks")

Features Real-time webcam-based inference

Adjustable confidence threshold

Lightweight and responsive UI using Streamlit

Model Info Framework: TensorFlow

Input: 63-dim vector (21 landmarks × 3)

Output: One-hot encoded sign labels

Label encoding: Stored in label_encoder.pkl

Try It 👉 Try the demo on Streamlit

👉 Try the demo on Hugging Face

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published