Recently, we proposed two spatiotemporal extensions of the local binary pattern operator for describing dynamic textures: Volume LBP (VLBP) and LBP from Three Orthogonal Planes (LBP-TOP). As our approach involves only local processing, we are allowed to take a more general view of dynamic texture recognition, extending it to specific dynamic events such as facial expressions or human actions. A block-based approach combining pixel-level, region-level and volume-level features was proposed for dealing with such nontraditional dynamic textures in which local information and its spatial locations should be taken into account.

In this talk we introduce our approach to dynamic texture analysis and survey the recent research in applying spatiotemporal LBPs to different motion analysis problems. Among the applications we consider are recognition and segmentation of dynamic textures, facial expression recognition, face and gender recognition, visual speech recognition, recognition of human actions and gait, and video texture synthesis.