Pattern recognition technologies for multimedia information processing

Advanced algorithms based on pattern recognition for segmentation, recognition, and indexing of image, video, speech and audio data have been investigated and developed to analyze multimedia information, recently. Pattern recognition algorithms generally aim to provide a reasonable answer for all possible inputs and to perform most likely matching of the inputs, taking into account their statistical variation. Theses algorithms have been widely used in many applications to recognize what people understand such as objects, faces, human actions, and so on. Although there has been a great progress in fields of image and speech processing, virtual and augmented reality, human-computer interaction, and so on, we still have the difficulties of implementing general models that are robust to real-world variations. In this special issue, our goal is to consolidate the recent research achievements that include diverse pattern recognition based information processing for multimedia applications.