This project focuses on developing an online hand gesture recognition system. Utilizing advanced deep learning techniques, it aims to accurately recognize and interpret hand gestures in real-time. The system employs a sophisticated CNN-Transformer architecture to process and analyze video data, making it capable of understanding complex hand gestures.
- Real-Time Gesture Recognition: Detects and interprets hand gestures in real-time.
- Advanced CNN-Transformer Model: Leverages the strengths of both CNNs for spatial feature extraction and Transformers for capturing temporal dynamics.
- Variable-Length Sequence Handling: Accommodates gestures of varying durations with effective padding and masking strategies.
- High Accuracy and Efficiency: Optimized for both high accuracy in gesture recognition and operational efficiency in online settings.
- Python 3.8 or above
- PyTorch
- Torchvision
- CUDA (for GPU acceleration)
- Other dependencies listed in
requirements.txt
Clone the repository and install the required dependencies:
git clone https://github.com/KenanKhauto/online-gesture-recognition-project.git
cd online-gesture-recognition-project
pip install -r requirements.txt