People Emotion Recognition


Raise your platform’s efficiency with the fast and powerful sentiment recognition

With this CNN for facial emotion recognition, you can save expenses for video analytics automation and raise the effectiveness of your business, customer satisfaction, and sales.

Product Overview

Applications The PER model can recognize on video streams, records, and images all types of people’s emotions according to Dr. Ekman:

  1. Happiness
  2. Sadness
  3. Fear
  4. Anger
  5. Surprise
  6. Contempt
  7. Disgust

It also recognizes all combinations of these emotions.


You can best use this model in retail marketing for:

  • analyzing customers’ reaction to your commodities and services
  • creating heat maps Installation and Use
  • assessing your marketing strategy

Installation and Use (Linux/MacOS)

1) download project / library

2) Install library `pip install *.whl` OR `python install`

How to use

  • CLI
  • Python interface
import cv2
from model import Model
import argparse
import sys

image = cv2.imread('../tests/surprised.jpg')
model = Model()
ret = model.process_sample(image)

# Text Output
image = cv2.putText(image, str(ret), (10, 50), cv2.FONT_HERSHEY_SIMPLEX, 
1, (255, 0, 0), 2, cv2.LINE_AA)

cv2.imshow('Emotions', image)
key = cv2.waitKey(5000) #pauses for 5 seconds before fetching next image


There are no reviews yet.

Write a review

Your email address will not be published. Required fields are marked *