Online learning is better with adequate emotion recognition
November 11, 2015 – A new dimension of online learning presents itself. PhD Kiavash Bahrainis developed and tested software that detects emotions on the face and the voice of students accurately and directly used for giving feedback.
The title of his thesis is “Real-Time Multi-Modal Emotion Recognition in e-learning ‘. Bahrainis receive his PhD on Friday, November 13th, 2015 at 16.00 at the Open University.
Defaults
Existing software for recognizing emotions has quite a few flaws. The reliability is limited, the software has problems with mustaches, beards, glasses, and drop shadows and the variety of emotions that can be recognized is small (sometimes only positive or negative). Often, expensive equipment is required to detect emotions and the outcomes are not always readily available as additional calculations are needed.
Better Software
Bahreini has managed to take great part of the problems road by combining a number of existing methods and expand. Its software is much more accurate and, moreover, is available in real-time. The software recognizes a wide spectrum of basic emotions (happiness, sadness, surprise, fear, disgust, anger, and neutral) on the face and in the voice of a student and has hardly affected by disturbances. A standard webcam and microphone suffice.
Improved academic performance
Bahrainis tested the software for a digital game-based training communication skills with students just had to show their emotions in check hold. He concludes that the software he has developed emotions and accurately detects that the computer feedback that follows leads to better learning outcomes. The feedback received will be also accepted without problems by the students and appreciated.
No comments:
Post a Comment