This project aims to develop a facial expression recognition system using PyTorch. The system classifies facial expressions into various categories using a convolutional neural network (CNN). The dataset for training and testing the model is sourced from Kaggle and includes multiple facial expressions.
The dataset used for this project is available on Kaggle. It contains images of faces categorized by different expressions such as happy, sad, angry, surprised, etc.
Dataset URL: Face Expression Recognition Dataset
To set up the environment and install necessary dependencies, follow these steps:
- Install Albumentations: This library is used for augmenting the dataset.
pip install -U git+https://github.com/albumentations-team/albumentations
- Install TIMM: PyTorch image models library by Ross Wightman.
pip install timm
- Upgrade OpenCV: Ensure you have the latest version of OpenCV which includes contributions.
pip install --upgrade opencv-contrib-python
The model architecture is based on convolutional neural networks (CNNs) designed for image classification tasks. It leverages pre-trained models from the TIMM library to enhance performance and reduce training time.