Strong presence at ACII and Ubicomp 2019

Researchers from the ROC-HCI lab attended two back-to-back conferences in the first half of September: ACII’19, held in Cambridge, and Ubicomp’19,  in London.

PhD student Md Kamrul Hasan presented a framework developed by his team that achieves 70% accuracy in recognizing when someone is remembering vs imagining. In contrast, humans can achieve only around 50% accuracy in such tasks.

Implications? Imagine algorithms adding more transparency to the human decisions on measuring credibility.

The second and third days of the ACII conference highlighted the works on social-network driven speaking-skills development by Baten et al., what computers can teach us about doctor-patient communication by Ali et al., visual cues for disrespectful conversation analysis by Samrose et al., and using facial features to contextualize linguistic analysis by Hasan et al.

Visiting Cambridge is incomplete without Punting and the delicious food!

At Ubicomp’19, PhD student Wasifur Rahman presented a work on recognizing PTSD using a wearable device. Taylan Sen and Raina Langevin presented on the remote assessment of Parkinson’s.

What would you do if asked to present your two years of work in 10 minutes during the final day of the conference right after lunch to a half asleep audience? How about this for a talk opening? 😀

PhD student Taylan Sen puts his head on ice-cold water to demonstrate the lower expressivity of facial expressions of individuals with Parkinson’s.