Loading...

Design and Implementation of Face Recognition System on Expressive Robot and Study of its Interactive Effects with Deaf Children

Rokhi, Zeynab | 2021

420 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 53888 (08)
  4. University: Sharif University of Technology
  5. Department: Mechanical Engineering
  6. Advisor(s): Meghdari, Ali; Alemi, Minoo; Taheri, Alireza
  7. Abstract:
  8. Social robots are a group of robots designed and built to interact with humans, so the ability to interact emotionally in these robots, especially in interaction with people with hearing and speech disorders, is essential because these people do not have an emotional state sound. Therefore, in this research, an emotional interaction system based on recognizing the user's emotional state from his facial state using deep learning algorithms as well as performing an appropriate emotional response on the robot's face has been designed and implemented on an expressive social robot. In order to design the emotional state recognition system, first, the best model was found on the large AffectNet emotional database. Then in order to improve the accuracy of this network using the transfer learning algorithm and the database collected by the robot camera, the parameters of the trained model were updated. After designing and implementing the mentioned system on Rasa social robot, to investigate the effect of the presence of social robot with emotional interaction system in speech therapy sessions, two target groups were selected, and one group under speech therapy sessions with the presence of a robot with mentioned capabilities and the other group in sessions without Robots are involved. For accurate and statistical analysis, items from the TOLD test were selected and taken from both groups before and after clinical sessions. These tests' statistical results show an improvement in quality and a significant increase in the effectiveness of speech therapy sessions with children with speech disorders through the presence of a social robot with the ability to interact emotionally in speech therapy sessions
  9. Keywords:
  10. Social Robotics ; Human Robot Interaction (HRI) ; Neural Networks ; Human Emotion Recognition ; Hearing-Impaired

 Digital Object List

 Bookmark

No TOC