Pukyong National University

PKNU received 'grand prize' at the 2021 AI Capstone Design Contest

- Advanced facial image composition research using deep learning

△ A picture of the winners. Seo Hye-Jin, Jo Soo-Hwa, Kim Chae-Gyu, Lee Tae-Kyeong, and Park Gyu-Bin from the left.

A research team from the Artificial Intelligence Research Center at Pukyong National University won the grand prize in the AI capstone design contest of the 2021 Fall Conference of the Korean Multimedia Society.

This research team, which consists of Lee Tae-Kyeong, Park Gyu-Bin, Jo Soo-Hwa, Seo Hye-Jin, an associate researcher Kim Cha-Gyu from the Artificial Intelligence Research Center at Pukyong National University and D. Eng. Song Bok-Deuk from the Electronics and Telecommunications Research Institute (ETRI) was recognized for the best performance by presenting a thesis 'A study on a method for facial composites that is strong face pose using a head pose vector' at this competition.

This study suggested a way to obtain a more improved facial composites result by applying the head pose vector value to the existing research related to facial composites in image processing.

The research team received good reviews for suggesting a method for a facial composites that is strong in facial pose by using deep learning, an artificial intelligence technique, to learn the movements and directions of the sides and eyes of the face using the head pose vector.

In the laboratory of an associate researcher Kim Chae-Gyu of the Artificial Intelligence Research Center at Pukyong National University, he is actively conducting researches in the field of artificial intelligence, including facial composites, and researching the development of technologies that use deep learning to convert images into sound and sound into images. <Pukyong Today>