https://github.com/msiddhu/emotion-detection

A emotion detection tool based on TensorFlow and OpenCV
https://github.com/msiddhu/emotion-detection

accuracy dataset emotion-detection faces numpy opencv python tensorflow tensorflow-models webcam-feed

Last synced: 3 months ago
JSON representation

A emotion detection tool based on TensorFlow and OpenCV

• Host: GitHub
• URL: https://github.com/msiddhu/emotion-detection
• Owner: msiddhu
• License: mit
• Created: 2020-01-29T07:09:47.000Z (almost 5 years ago)
• Default Branch: master
• Last Pushed: 2021-04-30T10:05:06.000Z (over 3 years ago)
• Last Synced: 2024-10-11T08:43:43.804Z (4 months ago)
• Topics: accuracy, dataset, emotion-detection, faces, numpy, opencv, python, tensorflow, tensorflow-models, webcam-feed
• Language: Python
• Homepage:
• Size: 21.4 MB
• Stars: 3
• Watchers: 2
• Forks: 2
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • Funding: .github/FUNDING.yml
  • License: LICENSE
  • Code of conduct: CODE_OF_CONDUCT.md

Awesome Lists containing this project

README

        
# Emotion-detection

## Introduction

This project aims to classify the emotion on a person's face into one of **seven categories**, using deep convolutional neural networks. This repository is an implementation of [this](https://github.com/msiddhu/Emotion-detection/blob/master/ResearchPaper.pdf) research paper. The model is trained on the **FER-2013** dataset which was published on International Conference on Machine Learning (ICML). Face images with **seven emotions** - angry, disgusted, fearful, happy, neutral, sad and surprised.

## Dependencies

* Python 3, [OpenCV 3 or 4](https://opencv.org/), [Tensorflow 1 or 2](https://www.tensorflow.org/)

* To install the required packages, run `pip install -r requirements.txt`.

## Usage

The repository is currently compatible with `tensorflow-2.0` and makes use of the Keras API using the `tensorflow.keras` library.

* pretrained model link-

* The folder structure is of the form:  

  Tensorflow:

  * data (folder)

  * `emotions.py` (file)

  * `haarcascade_frontalface_default.xml` (file)

  * `model.h5` (file)

* This implementation by default detects emotions on all faces in the webcam feed.

* With a simple 4-layer CNN, the test accuracy peaked at around 50 epochs at an accuracy of 63.2%.

![Accuracy plot](accuracy.png)

## Algorithm

* First, we use **haar cascade** to detect faces in each frame of the webcam feed.

* The region of image containing the face is resized to **48x48** and is passed as input to the CNN.

* The network outputs a list of **softmax scores** for the seven classes.

* The emotion with maximum score is displayed on the screen.