"Xiao Chu", a computer-generated woman on screen, attracts a visitor's attention at the 2021 East Lake International Artificial Intelligence Summit Forum on Dec 18. [Photo/WeChat account of Wuhan Future City]

Software functions, such as translating speeches into words, or translating one language into another, is nothing new for the majority of intelligent terminal users. But have you thought of transforming images into speeches, or converting speeches into images? That is the new breakthrough in cross-modal general intelligence made by the Wuhan AI Computing Center.

A virtual figure on the screen named "Xiao Chu" dressed in traditional Chinese costumes was created by the center, and is capable of communication with humans. She can form pictures based on descriptions and detect machine errors after listening to simple videos or speeches.

A staff member at the center explained how the AI technology is applied in everyday life. "For example, we can only use simple functions of common photo applications, such as broadening and extending eyebrows, but the intelligent model based on big data can tell us the best way to improve a photo."

The center has reportedly rendered its inclusive computing power services to more than 80 key enterprises, including Wuhan Yangtze Computing Technology Co and Wuhan Cobot Technology Co, and has participated in over 30 smart city infrastructure solutions, whose efficiency has increased by 40 percent due to these efforts.