Gesture Detection with TensorFlow and MobileNet

This project utilizes TensorFlow and MobileNet for real-time hand gesture detection. Initially, a custom Convolutional Neural Network (CNN) was implemented but later transitioned to MobileNet due to its efficiency and superior performance.

Features:

• Real-time Gesture Recognition: Detects static hand gestures with high accuracy.
• MobileNet Integration: Leverages a pre-trained MobileNet model for lightweight and fast inference.
• Custom Dataset Support: Allows integration of custom datasets for training and fine-tuning.
• Python-based Implementation: Simple and modular code, making it easy to modify and extend.

Use Cases:

• Sign language recognition.
• Gesture-based control systems.
• Human-computer interaction projects.

Getting Started:

1. Clone the repository.
2. Install dependencies: pip install -r requirements.txt.
3. Train your model or use the pre-trained MobileNet weights.
4. Run the real-time gesture detection script.

Challenges Overcome:

• Improved accuracy and performance by switching from a custom CNN to MobileNet.
• Optimized for lightweight applications and edge device compatibility.

Feel free to contribute, report issues, or suggest enhancements!