Mild cognitive impairment detection from facial video interviews by applying spatial-to-temporal attention module

Muath Alsuhaibani, Hiroko H. Dodge, Mohammad H. Mahoor

Abstract: Highlights•The semi-structured interviews from the I-CONECT study are obtained for this work.•The facial features of participants from the videos are extracted using CAE on a frame level.•We proposed STAM which integrated temporal facial features to detect MCI participants.•The interaction feature during the interview improved STAM prediction performance.