Apply for community grant: Academic project (gpu)
This project is a real-time micro-expression recognition application developed as part of an effort to democratize access to emotion AI research. It enables live facial emotion detection through the webcam, using a deep learning model integrated into a user-friendly web interface built with Streamlit and streamlit-webrtc.
Key features include:
- Live video inference directly from webcam input
- Accurate classification of micro-expressions such as anger, surprise, fear, and happiness
- Interactive and responsive UI, accessible from any modern browser
- Containerized deployment using Docker for reproducibility and scalability
- Open and modular codebase designed for experimentation, education, and community collaboration
Micro-expressions are fleeting facial movements that can reveal underlying emotions, often missed by the human eye. This tool offers a lightweight and accessible way to explore these expressions, with potential applications in behavioral research, human-computer interaction, mental health, and more.
By sharing this project openly, we aim to contribute to the open-science and open-source communities, making it easier for others to build on emotion recognition research without proprietary constraints.
Here is the Github link: https://github.com/homxxx/MicroExpressionRecognition/