Faculty-student project develops emotion-reading technology
Elizabeth Lamark/RIT Production Services
Facial recognition software, layered within interactive computer programs, can help people identify their emotions and learn to model healthy communication styles. Medical informatics graduate student Teraisa Chloros demonstrated the technology to local mental and behavioral health professionals.
Follow Susan Gawlowicz on Twitter
Follow RITNEWS on Twitter
Teraisa Chloros made faces at her laptop computer. Her mock expressions appeared on the overhead screen with labels identifying her emotions—happy, sad, surprised, angry, disgusted or scared—based on the position of her facial muscles.
Chloros ’11 (medical informatics) illustrated the FaceReader software by Noldus Information Technology to mental health professionals in March at a conference sponsored by the RIT College of Health Sciences and Technology and the Office of Research.
The software is one component of the interactive tool she developed with a faculty-student team led by Caroline Easton ’90 (biotechnology), RIT professor of forensic clinical psychology. Easton asked the team to create computer programs for her clients to practice what they learn in therapy. The multimedia tools will augment her clinical model for treating clients with co-occurring substance abuse and domestic/intimate partner violence.
Easton’s ideas depart from the standard treatment model that differentiates substance abusers and domestic violence offenders. Her integrative method of care connects the two mental and behavioral health problems and treats the individual, not the group. She is a proponent of adapting gaming technology to teach clients to reduce substance abuse and increase communication, anger management and healthy conflict-resolution skills.
“I think virtual reality tools are the future,” Easton says. “Human visualization, cyber tools and gaming are all going to be an important part, coupled with our best practice procedures.”
Ideas for computer-assisted therapy led Easton, then-director of Yale University Medical Center’s Forensic Drug Diversion Clinic, to tap her alma mater’s technical expertise. The RIT connection resulted in a co-op for Chloros at the clinic and the challenge of building the technical backbone to support Easton’s vision.
“My focus has been on the importance of reading emotion so clients can understand what their partners are expressing and what nonverbal behavior indicates,” Chloros says.
The website application and supporting databases represent Chloros’ capstone project for the master’s degree she will receive this year in medical informatics. The Web application resembles a gaming environment with educational videos, interactive modules with role-playing exercises and realistic vignettes, and self-assessments.
Therapists can use the videos as stand-alone educational resources or in conjunction with computer-assisted therapy, Chloros explains.
Integral to the application is Al-Virt, an empathetic avatar and customizable digital coach designed by Alan Gesek, who is graduating with his Master of Fine Arts in medical illustration. When fully developed, Al-Virt will interact with clients and model healthy communication skills.
“The next phase of the project would be to incorporate FaceReader into a program like Al-Virt,” Chloros says.
The multidisciplinary clinical research team working with Easton and Chloros includes RIT professors Richard Doolittle, Shaun Foster, Glenn Hinz, Stephen Jacobs and James Perkins; and students Valerie Altounian, Nirja Desai, Timothy Fitzgerald and Gesek.
“Human visualization and cyber tools are the future in mental and behavior health treatment and education,” Easton says. “RIT students are doing it now. They’re at the forefront.”
