Emotion Recognition- The Next Big Innovation

Shamna. V.H, M.P. Viswanathan

Abstract


Today, as we progress through each passing day the interaction, communication and connectionsbetween human and machinery have been more and more prevalent. Many types of software have been developed to take out human interventions to the extent possible and to make our life easy and convenient. An interesting thing to note here is that no level of automation or programming hasbeen successfully developed so far as a perfect solutionto understand the human sentiments or behaviors. However, the efforts in this direction are humongous. Continuous efforts are being made to make gadgets that can easily recognize signs of human beings and sentiments. Oncewe can have a mechanism to recognize human behavior in an enhanced way, there way do or experience everything now will move into a different stages of wonder.All our behavioralaspects such body language,creativity, perception, response-ability, facial expressions and emotions have huge impact in our daily life be it personal or

 

professional.Facial expression is one of the lively investigated subjects in Human Computer Interaction (HCI). Sensation can be well-known by tone of voice, expression, text and any kind oral communication. Facial expressions play a significant role in classifying feeling of human beings. Through this assignment, actual time emotion detection from facial representation is planned where 3 steps face recognition using

1) Haar cascade, 2.Skin tone origin using Active Shape Model (ASM) and 3. MultiSVclassifierare designed to classify five main human sentiments i.e. Anger, Disgust, Joy, Neutral and Sad. This is projected to be put into practice through a machine called MATLAB. The training handpicked datasetsof emotion detection has shown a standard result at 94% accuracy which is awesome.


Full Text:

PDF




Copyright (c) 2018 Edupedia Publications Pvt Ltd

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

 

All published Articles are Open Access at  https://journals.pen2print.org/index.php/ijr/ 


Paper submission: ijr@pen2print.org