INSTITUT FÜR PSYCHOLOGIE

 0P392 Advanced Perception

This course covers advanced knowledge about perceptual processes in interaction with technological media. The course starts with a treatment of quantitative modeling and will present basic human recognition models and models for information representation, as well as methods to develop those and estimate their parameters. Psychological models of representative visual and cognitive tasks, like reading, visual search, object and picture perception will be treated in detail. On the same quantitative basis the principles behind image coding are explained, with special emphasis on perceptual criteria in coding techniques that can produce optimal images with different degrees of compression. Current coding techniques as applied in JPEG and MPEG standards are presented and are elaborated in the exercises. Methods for subjective evaluation of perceptual image quality are discussed and related to the basic psychological models presented before, and to Thurstonian scaling.

In the area of audio, treated next, most attention is given to the composition and representation of speech, which is followed by techniques of audio coding and perception along the same lines as for video coding. Popular techniques of audio coding like MP3 and alternative methods will be discussed, and effects of sampling frequency and quantization levels will be illustrated in exercises. The perception of simultaneous visual and auditory stimuli is handled in the description of multimodal perception, where issues of synchrony and signal complexity are highlighted, as well as creation and perception of auditory scenes. Attention, more precisely attentional processing acting like a filter and as a focusing mechanism, is discussed with respect to its capability to concentrate on particular perceptual inputs and its vulnerability to distracting information. Some practical consequences of human perception are finally illustrated in examples of the concept of change blindness, which is closely related to the reliability of eye-witness testimony.

All material presented in the course is elaborated in the form of practical exercises that are made under supervision and with support of the teaching staff in separate (OGO) sessions.

Learning objectives:

• obtain the routine capability to develop, understand and evaluate models of human perception and cognition
• to understand the relation between human perception and methods of signal coding for visual and auditory information
• to obtain insight in the perceptual requirements applicable in a user-centered design approach to optimize current and novel technological developments in multi media application
• to get insight in the possibilities and limits of human perception in realistic situations.