A CNU research team won an international competition on emotion recognition and the prediction of changes in people's emotions.

The CNU Emotional Intelligence Joint Research Team (Department of Artificial Intelligence Convergence) consisting of Huynh Van Thong (Ph.D. candidate) and professors Kim Soo-hyung, Yang Hyeong-jeong, and Lee Gwi-sang participated in the International Competition on Evoked Expression from Video (EEV) Recognition, which took place on May 1. With a model titled Temporal Convolution Networks with Positional Encoding, it surpassed rival teams in the US, China, and Spain, and ranked first in prediction accuracy.

The EEV contest competitors use technology to predict the change in emotions people feel while watching a specific video, and the related technology is used for recommending videos according to individual preferences or creating advertisements or videos that appeal to people.  

The emotion recognition model of the CNU research team will be presented at the AUVi (Affective Understanding in Video) workshop session of the CVPR 2021 International Conference, which is one of the top-tier conferences in the field of computer vision and is scheduled to be held as an online conference from June 19 to 25. 

The CNU research team also took first place in the Prediction of the Intensity of Pain from Facial Expressions at the international contest for emotion recognition held online in November last year.