🖐 AI Hand Gesture Recognition System

A Deep Learning based Hand Gesture Recognition Web Application developed using TensorFlow, CNN, OpenCV, and Flask.

This project can recognize different hand gestures from uploaded images and display the predicted gesture along with an emoji representation.

---

📌 Project Overview

Hand Gesture Recognition is an important Computer Vision application that enables Human-Computer Interaction (HCI).

This system uses a Convolutional Neural Network (CNN) trained on the LeapGestRecog dataset to classify hand gestures into 10 different categories.

The trained model is integrated with a Flask web application where users can upload gesture images and receive instant predictions.

---

🎯 Features

• Upload hand gesture images
• Real-time gesture prediction
• CNN-based deep learning model
• Flask web application
• Image preview
• Emoji-based gesture display
• Modern responsive UI
• Trained TensorFlow model

---

🛠 Technologies Used

• Python
• TensorFlow / Keras
• OpenCV
• NumPy
• Scikit-Learn
• Flask
• HTML
• CSS

---

📂 Dataset

Dataset Used:

LeapGestRecog Dataset

https://www.kaggle.com/datasets/gti-upm/leapgestrecog

Dataset contains 10 gesture classes:

Label	Gesture
01	Palm
02	L
03	Fist
04	Fist Moved
05	Thumb
06	Index
07	OK
08	Palm Moved
09	C
10	Down

---

🧠 Model Architecture

CNN Architecture:

Input Layer (64x64x3)

↓ Conv2D (32 Filters)

↓ MaxPooling2D

↓ Conv2D (64 Filters)

↓ MaxPooling2D

↓ Flatten

↓ Dense (128)

↓ Output Layer (10 Classes)

---

📁 Project Structure

PRODIGY_ML_04/
│
├── data/
│   └── leapGestRecog/
│
├── model/
│   └── hand_gesture_model.h5
│
├── static/
│   ├── style.css
│   └── uploads/
│
├── templates/
│   └── index.html
│
├── flask-app.py
├── train_model.py
├── requirements.txt
├── README.md
└── screenshots/

---

⚙️ Installation

1 Clone Repository

git clone https://github.com/yourusername/PRODIGY_ML_04.git

2 Navigate to Project Folder

cd PRODIGY_ML_04

3 Install Dependencies

pip install -r requirements.txt

---

🚀 Train Model

Run:

python train_model.py

After training, model will be saved as:

model/hand_gesture_model.h5

---

🌐 Run Flask Application

python flask-app.py

Open browser:

http://127.0.0.1:5000

---

📊 Results

• Training Accuracy: 99.9%
• Validation Accuracy: 100%
• Gesture Classes: 10

---

📷 Screenshots

Home Page

(Add Screenshot Here)

Prediction Result

(Add Screenshot Here)

---

🔮 Future Improvements

• Real-time Webcam Recognition
• Live Gesture Tracking
• Mobile Responsive Design
• Gesture History Tracking
• Voice Feedback System

---

👨‍💻 Author

Developed as part of the Prodigy InfoTech Machine Learning Internship.

Task-04

Hand Gesture Recognition using CNN, TensorFlow, OpenCV and Flask.

---

📜 License

This project is for educational and internship purposes.