The processing of sensory information from the eye, ear, or tactile system into its "internal representation" was investigated, analysed and modelled with a multidisciplinary approach. Both the structure and function of certain subsystems were considered (e.g., neural networks in the retina) as well as their respective cooperative performance in well-defined psychophysical tasks (such as, e.g., loudness perception in normal and hearing-impaired listeners). Special emphasis was placed on object perception and the physiological basis of object representation which is presumably characterized by similarities and interactions across sensory systems. Since the research methods employed cover molecular genetics, single-cell electrophysiology, human and animal psychophysics, EEG recording, fMRT, and numerical modelling, they provided a broad methodological frame for high-level Ph.D. projects. Parts of the results were used for clinical and technical applications such as improved interpretation of functional images in neurosensory tasks or improved man-machine-communication by employing auditory-model based processing techniques. 
 

How do our sensory systems transform physical stimuli into the corresponding neural representations and how can we employ this knowledge for selected applications? This and related questions created the central topic of the international graduate school (InterGK) for Neurosensory Science, Systems and Applications. As opposed to classical neuroscience, the InterGK neither focused on receptor biophysics nor on neuronal activity of circumscribed parts of the central nervous system alone. Instead, it tried to bridge the gap between the first transformation stages of the physical stimulus and the complex central processing of stimulus features and objects as well as the gap between basic neuroscience and applied physics and engineering. This was achieved by analyzing and modelling the transformation from the physical stimulus into its "internal representation" and by exploiting the resulting knowledge both to develop practical applications and to test the underlying assumptions and models. Hence, a variety of methods, sensory modalities, neurosensory systems and research questions has to be covered in the InterGK. 
 

As central common research topics we identified the following three areas 
 

• a. Dynamics of sensory systems
• b. Object perception
• c. Applications

All three areas included experiments with various methods as well as modelling at different levels of detailedness (i.e., from the behaviour of single neurons or neural assemblies to models of "effective" processing in the nervous system). While area a) considered primarily bottom-up processing schemes that are primarily driven by the physical stimulus, area b) primarily considered top-down processing schemes and their interaction with bottom-up processing in order to shape objects in the internal representation. Hence, area a) was primarily concerned with the receptor and the consequences of stimulus dynamic range and stimulus time structure on the internal representation of the stimulus. Conversely, area b) was primarily concerned with the central processing of sensory information. Finally, area c) utilized the input from both other areas to advance different applications. All three areas exhibited a considerable overlap with the four main research areas that were used to structure the workprogram of the InterGK in the previous reporting period, i.e., adaptation, object perception, imaging and application.