Yun Ho of University of Chicago

Wearable interfaces are bringing computer assistance into the real world. For instance, smart glasses understand the surrounding context in which users encounter difficulties. These advances in sensing and reasoning make it possible to go beyond the traditional forms in which computers assist users (e.g., textually, visually), and instead, assist physically. The key aim of physical assistance is to impart tacit knowledge to the user, the “know-how” of performing a specific task (e.g., the key gesture that releases the locking mechanism on a pill bottle).

Currently, computer interfaces represent and communicate information through symbols, such as visuals, text, or speech, leaving out tacit knowledge, which is encoded in gestures or forces. In this talk, I will present prototype systems that employ body actuation techniques (e.g., electrical muscle stimulation) to enable computational intelligence to communicate with users through body movements.
I will share how we integrated multimodal AI with muscle stimulation to generate step-by-step movement instructions, what embodied intelligence can offer users (e.g., learning a sequence of movements or inspiring creative movements), and how understanding users’ interpretation of machine-induced movements can help us design better physical assistance. These instances allow us to explore the benefits and limitations of embodiment in interfaces, and how embodied and symbolic interfaces complement each other.

Biography:

Yun Ho (https://www.yunho.org/) is a Ph.D. student in computer science at the University of Chicago advised by Prof. Pedro Lopes. She explores how people make sense of and build relationships with physical assistance systems that communicate with users by means of movements & gestures (e.g., via their sense of proprioception). One instance of this is her system Embodied AI (https://embodied-ai.tech), which assists users in performing new physical tasks through generated electrical muscle stimulation instructions adjusted to the user’s needs & context. You can find her work published at ACM CHI (the premier publication venue in human-computer interaction), for which she has received two CHI best paper awards.