A joint project of the Graduate School, Peabody College, and the Jean & Alexander Heard Library

Title page for ETD etd-09132016-150723

Type of Document Dissertation
Author Zheng, Zhi
URN etd-09132016-150723
Title Machine-assisted Technologies for Young Children with Autism Spectrum Disorder: Novel Platforms for Early Detection and Intervention
Degree PhD
Department Electrical Engineering
Advisory Committee
Advisor Name Title
Nilanjan Sarkar Committee Chair
Amy S. Weitlauf Committee Member
D. Mitchell Wilkes Committee Member
Gabor Karsai Committee Member
Robert J. Webster III Committee Member
Zachary E. Warren Committee Member
  • human-robot interaction
  • children with ASD
  • human-computer interaction
  • Autism Spectrum Disorder
Date of Defense 2016-09-08
Availability unrestricted
Autism Spectrum Disorder (ASD) is a neuro-developmental disorder with a high prevalence rate of 1 in 68 children in the U.S. Human-Machine Interaction (HMI) is being continuously explored as a potential efficacious intervention tool for young children with ASD. While initial studies are encouraging, several challenges exist, including: 1) how to target the core deficits of ASD using technologies; 2) how to make the systems adaptive based on children’s real-time response; 3) how to detect interaction cues non-invasively; and 4) how to validate skill generalization from machine-assisted intervention to human-human interaction. This dissertation addresses these challenges by designing intelligent systems and user studies targeting three core deficit areas of ASD, which are imitation, social orienting, and joint attention impairments.

First, we designed two autonomous robotic systems, named RISIA1 and RISIA2, to teach imitation skills to children with ASD. In RISIA1, we developed a novel non-invasive gesture detection method that allowed the robot to detect even partially completed gestures and give feedback to children in real-time. User studies showed that the children with ASD paid more attention to the robot than a human therapist and performed significantly better. Then, we expanded our gesture detection algorithm to include more complex gestures in RISIA2. Second, an autonomous computer-based system, named ASOTS, was developed to teach social orienting skills to the children with ASD. This system provides adaptive social orienting prompts through a novel attention attracting mechanism and non-invasive real-time gaze detection. User study showed that this system attracted and accurately detected the participants’ attention, and stimulated response to name calling behavior with high success rate. Finally, we designed a fully autonomous robot-mediated joint attention intervention system named Norris. This system is embedded with a novel large range, unobtrusive gaze tracking method and an adaptive prompting hierarchy. Longitudinal user studies indicated improved within-system performance as well as improved social communication skills in human-human interaction after robot-mediated intervention.

  Filename       Size       Approximate Download Time (Hours:Minutes:Seconds) 
 28.8 Modem   56K Modem   ISDN (64 Kb)   ISDN (128 Kb)   Higher-speed Access 
  Zheng.pdf 3.38 Mb 00:15:38 00:08:02 00:07:02 00:03:31 00:00:18

Browse All Available ETDs by ( Author | Department )

If you have more questions or technical problems, please Contact LITS.