BikeGesture: User Elicitation and Performance of Micro Hand Gesture as Input for Cycling

by | Feb 5, 2017

Authors: Yanke Tan, Sang Ho Yoon, Karthik Ramani
In Proceedings of the 2017 CHI Conference Extended Abstracts on Human Factors in Computing Systems (CHI EA' 17: 2147-2154), Denver, CO, May 6-11, 2017

Abstract: The use of hand gestures has a potential as a promising input metaphor. Wearables like smart textile and data gloves can provide hand gesture recognition to potentially replace, augment or improve existing input methods. Although recent bikes provide advanced functions with electro-mechanical components, the input metaphor still relies on mechanical switches or levers. In this paper, we investigate the acceptance and performance of using hand gesture during cycling. Through an observational study with 16 users, we devised a taxonomy of hand gestures. Users prefer subtle micro hand gestures to ensure safe cycling while maintaining a flexible controllability. We also implemented a wearable prototype that recognizes these gestures. In our evaluation, the prototype shows an average of 92 % accuracy while showing similar response time to existing mechanical inputs.

Sang Ho Yoon

Sang Ho Yoon

Sang Ho Yoon is currently working at Microsoft, Seattle, WA. He received his PhD at Purdue University and his B.S & M.S degrees from Carnegie Mellon University in 2008 with major in Mechanical Engineering and minor in Robotics. He worked at Research Department in LG Display & LG Electronics for 5 years. There, he involved in product development for consumer electronics as well as the futuristic products including 'Transparent & Public Display', 'Assistive/Rehabilitation Robot', and 'Smart Car User Interface'. He is particularly interested in applying novel sensing techniques to bring the new forms of input metaphor for Human-computer interaction. Areas of interest include wearable/tangible interface, sensing techniques & fabrication, and novel input device. Currently, his research aims at combining the state-of-art machine learning approaches with novel sensing technique to better support natural human-computer interaction. [Personal Website][LinkedIn]