This repository contains a pytorch implementation of an emotion classification algorithm, which is basically an image classification task. Following is the result of training efficient-net-b0 on Microsoft FER(Facial Expression Recognition) Dataset.
download requirements using conda
Download this dataset if you want to train your model from scratch
make sure you have git and conda installed on your system, then follow this steps
git clone https://github.com/yMayanand/Emotion-Recognition.git
cd Emotion-Recognition
conda env create -f envs/test_env.yml
conda activate test-emotion
python video_cam.py
you can also try my project directly from here which is live on huggingface spaces.