Qualcomm Institute, University of California San Diego
DONATE
Menu
Qualcomm Institute, University of California San Diego
Figuring out how to accommodate individuals with social communication differences or neurocognitive differences is always a two-way street between individuals and society. Our overall project seeks to find approaches for both sides of the street: the individuals with autism and the employers who hire them. The first three research projects discussed below explore ways to help the neurodiverse be more successful in their social communication in workplace environments (including in interviews), while the ethics group is exploring ways society can evolve to make the recruitment process safer for people on the spectrum.
With an interviewer avatar asking a series of questions pertaining to a tech job, our VR application allows a user to practice a mock job interview. The user’s gaze is tracked both when the avatar is speaking and when the user is speaking, to evaluate the social modulation of gaze. With two different interviews, each about half an hour in length, the app allows immersive situational practice of eye contact and of answering interview questions. Read more in our 2022 paper in the IEEE Transactions on Neural Systems and Rehabilitation Engineering, and our 2023 ISMAR conference paper.
Orienting towards people can be a way to show attention or include everyone in a conversation, but it can be a challenging part of social communication. Using LiDAR sensors to estimate body and head orientation, we found differences between neurotypical and autistic individuals. The system could be useful for behavioral feedback and situational practice. Early work is described in this 2021 Eusipco paper, and more recent work improves the estimation accuracy and extends the work to triadic conversations.
Head gestures such as nodding indicate active listening and show interest in a conversation. We used augmented reality glasses to measure an individual’s head movement during conversation, and created a machine learning system which could use the motion information to predict an overall conversational score. The system’s scores had a high correlation with the engagement scores provided by human raters.
Personality tests used for pre-employment screening are a large and growing business. Individuals on the autism spectrum experience a high unemployment rate, and often can be screened out by pre-employment personality tests that are not necessarily considering job-relevant skills. Looking over the history and current usage of such tests, this research examines the ethics of their use, impact on the neurodiverse, and policy recommendations. This work is part of a broader study aimed at building a more “ASD safe” recruitment process
Neurodiversity in Tech, Qualcomm Institute
University of California San Diego | Copyright 2022 – All Rights Reserved