A Vb Two

In the A-VB Two task, we explore low-dimensional affective space based on the circumplex model of affect, where participants predict the arousal and valence values (1=unpleasant/calm, 5=neutral, 9=pleasant/excited) as a regression task, and report the average Concordance Correlation Coefficient (CCC) and Pearson correlation coefficient on both dimensions. This task aims to enhance the accuracy and robustness of affect recognition, which holds significant application value in human-computer interaction, mental health assessment, and other fields.