Side Projects


“I’ve Got a Secret:” Trustworthiness of AI with Known Weaknesses

Artificial intelligence (AI) systems are becoming increasingly ubiquitous and often make high-impact decisions regarding human health and safety. However, such systems are known to lack explainability and make unexpected decisions, thus lowering user trust in many applications. Therefore, this project investigates how communicating a system weakness affects user trust.

Human-Centered Input Recognition Algorithms: Examining the Effect of Familiarity with Zhuyin Symbols on Recognition Accuracy using the $P Recognizer

The objectives of this graduate course project are two-fold: (1) to examine the performance of the $P recognizer on a new dataset called the Zhuyin Multistroke Gestures (ZMG) dataset and (2) to investigate whether familiarity with the Zhuyin symbols may affect the recognition accuracy.

Digital Health: Active and Healthy Living