This notebook is designed to train a deep learning model for face emotion recognition. It uses TensorFlow and is run in a Google Colab environment.
Face Emotion Model Training Notebook This notebook is designed to train a deep learning model for face emotion recognition. It uses TensorFlow and is run in a Google Colab environment. The workflow involves:
- Google Drive Integration: The notebook mounts Google Drive for loading data and saving model checkpoints.
- TensorFlow Setup: Specific versions of TensorFlow and Keras are installed to ensure compatibility with the code.
- TensorBoard Integration: TensorBoard is set up to monitor training metrics during the model training process.
- Kaggle Integration: The notebook connects to Kaggle for accessing datasets, using stored user credentials.
- Data Preparation: The AffectNet dataset is extracted and prepared for training.
The notebook progresses through model definition, training, and evaluation phases, with detailed tracking of performance using TensorBoard.