《TAIPEI TIMES》 Professor touts EmoMask idea

By Jake Chung / Staff writer, with CNA

A professor has turned to artificial intelligence （AI） in a bid to help people express themselves from behind masks amid the COVID-19 pandemic.

請繼續往下閱讀...

A project to create EmoMasks headed by National Tsing Hua University associate professor Pao Sheng-ying （包盛盈） uses AI to detect and express emotional states, hoping to break down the barrier to such behavior created by a government mask mandate.

Standard masks have become the new norm in Taiwan and the world since the outbreak of COVID-19, and while people in Taiwan are more accepting of them, wearing a mask abroad still projects aloofness or indifference, Pao said on Sunday.

Citing unspecified research, she said that non-verbal communication, especially among humans, largely relies on the muscles of the mouth and nose, which are covered by masks.

After masks were mandated in Taiwan in almost all settings after an outbreak of COVID-19 in May, the rules — along with social distancing orders — she wondered whether there was a way to maintain a semblance of normal interaction while sticking to the rules, Pao said.

After discussing her thoughts with National Taiwan University Department of Mechanical Engineering student Liang Shun-hsun （梁舜勛）, the idea resulted in the creation of the EmoMask, Pao added.

The EmoMask is an interactive, electronic device that gauges emotions with speech recognition and sensors, and outputs feelings, semantic animations or other interactive messages on a display panel, she said.

EmoMasks might be useful in factories, hospitals and tertiary service industries, she said.

However, development is still in its preliminary phase and the specialized mask has not yet been developed for commercial use, Pao said.

The design has allowed her to keep in touch with her academic roots — having obtained a doctorate at the MIT Media Lab, a research laboratory at the Massachusetts Institute of Technology — and is another step on the path to integrating technology and art, she said, adding that her previous work, Light Byte, was a similar attempt.

Light Byte used a real-time heat-sensing algorithm to manipulate wood panels on an installation to allow light into a room in patterns based on user input — typically crude drawings.

新聞來源：TAIPEI TIMES