Skip to content

FER model uses Convolutional Neural Network algorithm trained on FER2013 dataset to recognize five universal emotions of the Facial Action Coding System. It can detect multiple faces with exceptionally outstanding results. (IEEE Published)

Notifications You must be signed in to change notification settings

Shubs044/Facial-Expression-Recognition-Model

Repository files navigation

Facial Expression Recognition with Deep Learning

This project aims to develop a facial expression recognition system using Convolutional Neural Network (CNN) algorithm. The system is designed to recognize five universal emotions - happy, sad, neutral, surprise and angry - using the Facial Action Coding System (FACS).

To achieve this, we trained the system with the FER2013 dataset and implemented several pooling layers to extract features from the images. We also used haarcascade classifier to identify the presence of faces in the frame and detect the expressions of multiple faces at the same time.

The training accuracy obtained by the system is 73.12%. In real-time experiments, the system performed exceptionally well in recognizing the emotions of individuals.

To run the program, please ensure that all necessary libraries and dependencies are installed. Refer to the requirements.txt file for more information.

About

FER model uses Convolutional Neural Network algorithm trained on FER2013 dataset to recognize five universal emotions of the Facial Action Coding System. It can detect multiple faces with exceptionally outstanding results. (IEEE Published)

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages