Cascade of Tasks for facial expression analysis super()

Automatic facial action unit (AU) detection from video is a long-standing problem in facial expression analysis. Existing work typically poses AU detection as a classification problem between frames or segments of positive and negative examples, and emphasizes the use of different features or classifiers. In this paper, we propose a novel AU event detection method, Cascade of Tasks (CoT), which combines the use of different tasks (i.e., frame-level detection, segment-level detection and transition detection). We train CoT sequentially embracing diversity to ensure robustness and generalization to unseen data. Unlike conventional frame-based metrics that evaluate frames independently, we propose a new event-based metric to evaluate detection performance at the event-level. The event-based metric measures the ratio of correctly detected AU events instead of frames. We show how the CoT method consistently outperforms state-of-the-art approaches in both frame-based and event-based metrics, across four datasets that differ in complexity: CK+, FERA, RU-FACS and GFT.