/Facial-Expression-Detection-Using-Double-Bounded-Rough-Neutrosophic-Sets-

The most predominant non-verbal communication method used to understand the mentality and the emotion of a human being primarily revolves around facial expressions, gestures, and body language . As a result, biometrics like facial recognition will be critical in driving a conversational user experience. Facial recognition is becoming a standard security feature on devices like smartphones, and the technology is increasingly capable of recognizing emotion and sentiment through human’s facial expressions. These expressions also help us differentiate between genuine emotions and, for example, sarcasm.

Primary LanguageJupyter Notebook

Facial-Expression-Detection-Using-Double-Bounded-Rough-Neutrosophic-Sets-

The most predominant non-verbal communication method used to understand the mentality and the emotion of a human being primarily revolves around facial expressions, gestures, and body language . As a result, biometrics like facial recognition will be critical in driving a conversational user experience. Facial recognition is becoming a standard security feature on devices like smartphones, and the technology is increasingly capable of recognizing emotion and sentiment through human’s facial expressions. These expressions also help us differentiate between genuine emotions and, for example, sarcasm.

This project mainly focuses on a hybrid intelligent structure called “rough Neutrosophic sets” and also introducing “double bounded rough Neutrosophic sets” which are used for the important facial expression recognition. The significance of introducing this hybrid set structures is that the computational techniques based on any one of these structures alone will not always yield the best results but a fusion of two or more of them can often give better results.


The Outcomes of the project are given below:

  1. Detcting the facial expression of a person using Real Time Data
  2. Differentiating between Genuine and Fake Smiles
  3. Predicting if person might be Depressed
  4. Determing the Degree of Closeness to a particular Attribute/Expression
  5. With the onset of the corona virus pandemic, most people will choose to wear masks on a regular basis. By using Attribute Based Double Bounded Rough Neutrosophic Sets the person’s true expression can be detected just by using the feature points in and around the eyes.
  6. To check if a person is Faking an Expression or trying to Cheat during an Examination.

The repository contains the following files:

  1. Images_Universal.zip : Consists of 200 images which constitutes the Universal Set.
  2. facial_keypointf.csv : The csv file contains the Paramter Values for each of the 200 Images.
  3. Facial_Expression_Detection_Without_MASK.ipynb : It contains the code for detecting the facial expression when the person is not wearing a mask
  4. Facial_Expression_Detection_With_MASK.ipynb : It contains the code for detecting the facial expression when the person is wearing a mask
  5. requirements.txt : Consists of the Required Tools.