https://github.com/ahasannn/real-time-expression-detector

A Real Time Expression Detector using Python and Machine Learning
https://github.com/ahasannn/real-time-expression-detector

csv expression-detection-from-cam gradient-boosting logistic-regression machine-learning-algorithms mediapipe numpy opencv pandas python random-forest ridge-classifier scikit-learn

Last synced: 6 days ago
JSON representation

A Real Time Expression Detector using Python and Machine Learning

• Host: GitHub
• URL: https://github.com/ahasannn/real-time-expression-detector
• Owner: Ahasannn
• License: mit
• Created: 2022-11-19T11:33:51.000Z (about 2 years ago)
• Default Branch: master
• Last Pushed: 2022-11-19T17:48:04.000Z (about 2 years ago)
• Last Synced: 2024-11-09T20:44:01.066Z (2 months ago)
• Topics: csv, expression-detection-from-cam, gradient-boosting, logistic-regression, machine-learning-algorithms, mediapipe, numpy, opencv, pandas, python, random-forest, ridge-classifier, scikit-learn
• Language: Jupyter Notebook
• Homepage:
• Size: 8.17 MB
• Stars: 1
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Real Time Expression Detector

It is a realtime expression detector using MediaPipe, Python  & Machine learning algorithms.



## Brief Description

The realtime expression detector is designed using  Media Pipe and Python. Users can be able to leverage

their webcam to decode what their body language says at a point in time. So, specifically in

order to do this, the detector is leveraging pre-trained data as well as a custom machine learning model

to be able to take the landmarks from user's face as well as different poses from their body.




## Outline of Methodology 

![Working_Model](Images/Working_Model.png)



## Dataset

OpenCV and CSV are used to capture the landmarks related to classified facial and

gesturing expressions. The collected data are stored in the spreadsheet which are used to train the

machine to assume the real time detected poses. The more we collect landmarks for different

body and face gestures, the more we get the accurate prediction results.




## Machine Learning Models

I have used the following 4 machine learning models for expression detection.

* Logistic Regression

* Ridge Classifier

* Random Forest

* Gradient Boosting

### Comparisons Between Models

![Happiness_Detection](Images/Happiness_Detection.png)

![Sad_Detection](Images/Sad_Detection.png)

![Victory_Detection](Images/Victory_Detection.png)

## Results

![Results](Images/Results.png)

## Required Resources

* Language: Python

* Platform: Jupyter Notebook IDE

* Packages: MediaPipe, cv2, csv, os, numpy, pandas, sklearn, train_test_split, pickle