Hand Gesture Recognition

This repository contains code for training a machine learning model to recognize hand gestures from videos and a Python script to predict hand gestures in real-time using a webcam.

Getting Started

Prerequisites

• Python 3.x
• Required Python libraries: mediapipe, joblib, Flask, scikit-learn

Installation

1. Clone this repository to your local machine:

   git clone https://github.com/ArjunVit/hand-gesture-recognition.git
   

2. Install the required Python libraries:

   pip install mediapipe joblib Flask scikit-learn
   

Usage

Training the Model

1. Place your training videos in the videos folder. Each class should have its own folder containing video clips of that class.
2. Run the main.ipynb notebook to train the machine learning model. This notebook will preprocess the video data and train the model using scikit-learn.

Real-time Prediction

1. Run the predict.py script to start the Flask server:

   python predict.py
   

2. Open a web browser and navigate to http://localhost:5000 to view the real-time prediction dashboard.
3. The webcam feed will be displayed along with the detected hand landmarks. The corresponding buttons will be highlighted based on the predicted hand gesture.