Skip to Main Content
Summary form only given. "Multimodal" refers to the different senses (visual, audio, tactile, etc.) used in human-computer interface. "Multimedia" refers to the different ways of representing information (text, graphics, audio, images, video, etc.). A signal processing, analysis, or understanding task is called multimedia/multimodal if it involves two or more modalities or media interacting in nontrivial ways. We shall give an array of examples of multimedia/multimodal signal processing, analysis, and understanding, including audio/visual speech recognition and audio/visual emotion recognition. A stable and robust facial movement tracking algorithm is presented which is used in both tasks.