Facial Expression Recognition (FER2013)
AI/ML • 2023
Role: ML Researcher
CNN/MobileNetV2 trained on FER2013 dataset.
Features
- 68.2% test accuracy on 50k+ facial images
- Data preprocessing: resizing, normalization, augmentation
- Trained MobileNetV2 for robust real-time FER
- Integrated with OpenCV for deployment
Technologies
PythonTensorFlowCNNMobileNetV2OpenCV