Micro-expression spotting: A new benchmark

Micro-expressions (MEs) are brief and involuntary facial expressions that occur when people trying to hide their true feelings or conceal emotions. Based on psychology research, MEs play an important role in understanding genuine emotions, which leads many potential applications. Therefore, ME analysis has become attractive topic for various research areas, such as psychology, law enforcement, psychotherapy. In the computer vision field, study of can be divided into two main tasks, spotting recognition, used identify positions videos determine emotion category detected MEs, respectively. Recently, although much been done, no fully automatic system analyzing yet constructed a practical level reasons: most only focuses recognition part, while abandoning task; current public datasets not challenging enough support developing robust algorithm. The contributions this paper threefold: (1) we introduce extension SMIC-E database, namely SMIC-E-Long is new benchmark spotting; (2) suggest evaluation protocol standardizes comparison techniques; (3) extensive experiments with handcrafted deep learning-based approaches database performed baseline evaluation.