We all have experienced missing important sound signals (targets) although these were perfectly represented by the ear. This “blindness” for detecting targets in complex acoustic scenes has been termed informational masking. It is related to the statistical variation of the acoustic input and the complexity of acoustic scenes. How informational masking can be ameliorated by combining localization cues with other cues will be investigated in a combined behavioral and neurophysiological study in Mongolian gerbils. The under-standing of the mechanisms underlying informational masking will guide the enhancement of signals for human perception making them more salient.
PD Dr. Frank Ohl
The processing of acoustic information in auditory learning is known to involve discernible aspects of bottom-up and top-down processing. The project investigates bottom-up and top-down processes and their inter-action during processing of frequency-modulated sounds in auditory cortex. Using a combination of behavioral, electrophysiological and pharmacological approaches, as well as techniques and experimental paradigms developed in the first two funding periods, the neuronal circuits that implement the discernible processes are identified and experimentally modulated in a controlled fashion.
Dr. André Brechmann, PD Dr. Michael Brosch, Prof. Dr. Henning Scheich
(), (), ()
Our previous research indicates that the auditory system makes predictions about future stimuli to improve auditory processing and perception and that the auditory cortex is endowed with relevant prediction mechanisms, namely entrainment of brain oscillations with periodic stimulation and ramping activity. The 3rd funding period aims to uncover additional prediction mechanisms in auditory and prefrontal cortex of humans and monkeys that hold in situations in which the timing of upcoming stimuli is ill defined and in which prediction has to rely on other sound features. Special emphasis will be on finding brain mechanisms directly reflecting predictions, in addition to mechanisms reflecting the comparison of the prediction and the actual stimuli.
Prof. Dr. Georg Klump
The European starling provides a suitable model for investigating the neuronal mechanisms underlying auditory stream segregation since it perceives auditory streams in a similar way as humans do and its auditory forebrain neurons show response patterns that correlate well with human and starling perception. Our project aims at understanding the selection of cues in segregating sources in realistic scenes offering multiple cues at the same time either from the auditory modality alone or from both the visual and auditory modality. The results will reveal how the brain can be successful in analyzing the sounds from individual sources in a complex acoustic scene.
PD Dr. Peter Heil
The major goal of the project is to understand temporal processing mechanisms underlying the detection and recognition of sounds with different temporal envelopes. We developed a probabilistic, physiologically plausible model which successfully accounts for absolute thresholds of stimuli of different temporal envelopes. Here, we plan to explore its predictions for the shape of psychometric functions, for the detection of complex stimuli consisting of multiple spectral components relative to those of individual components, and for the discrimination of onset transients of supra-threshold sounds and its physiological basis. If successful, our model should replace the common acoustic energy-based models.
Prof. Dr. Thomas Münte
Clinic for Neurology, University of Lübeck
Dr. Tömme Noesselt, Prof. Dr. Hans-Joachim Heinze
Our project investigates the neural basis and temporal dynamics of cue-selection in audiovisual contexts in humans. In the last two funding periods we investigated the neural processes of temporal AV processing and top-down influences on AV-integration. We now extend our research and focus on the section of cues, which modulate AV-integration. In particular, we will investigate cue selection and the built-up of expectancies in regular and irregular temporal sequences and in behaviorally relevant looming stimuli. The results of our experiments will significantly broaden our understanding of the interactions of expectancies, behavioral relevance and action planning with audiovisual integration.
Prof. Dr. Christoph Herrmann
This project will investigate correlations between parameters of temporal resolution of the human auditory system and electrophysiologic brain responses as well as the level of cortical excitation. We expect that oscillatory processes and cortical inhibition will show such correlations. After identifying physiological processes, they shall be speeded up via repetitive transcranial electric stimulation. We hypothesize that a speedup of oscillatory brain responses will in turn improve auditory temporal resolution. Due to existing animal experiments, we expect a similar improvement for increased inhibition of auditory cortex.
Dr. Ulrike Langemann ()
Animal Physiology & Behavior Group, Dept. of Neuroscience, University of Oldenburg
Dr. Eike Budinger ()
Dept. Auditory Learning and Speech, Leibniz Institute for Neurobiology Magdeburg
Prof. Dr. Christine Köppl ()
This project addresses the question of how sound location is processed and represented through the minute interaural time difference with which sounds reach the two ears. We will take an evolutionary approach, studying select avian and mammalian species with different evolutionary histories and different behavioural ecologies. The aim is to clarify what alternative neural mechanisms exist at the level of the first binaural comparison in the brainstem and what determines their selection. The outcome will ultimately enable us to infer how the human brain localises sound. More generally, it will further our understanding about how key constraints shape the evolution of neural circuits.
Prof. Dr. Stefan Debener
The auditory temporal attending theory predicts that temporal regularities entrain attentional oscillations and thereby facilitate processing of sounds that adhere to a previously learned regularity. Using a pitch comparison task we test the hypothesis that the underlying neurophysiological mechanism is the entrainment of the phase of ongoing cortical oscillations. By means of human high-density EEG studies we will address whether cross-modal sensory entrainment explains the influence of regular visual and tactile stimuli on auditory cortex oscillations and task performance. We will also study whether a similar effect can by induced by transcrainal electrical stimulation, and what the role of primary sensory encoding on cross-modal entrainment is.
Prof. Dr. Jochem Rieger
In natural settings, speech sounds are accompanied by visual cues from the speaker. Such visual cues strongly modulate speech comprehension. We propose to characterize the mechanisms of bottom-up and top-down multimodal integration of visual cues in the neural encoding of spectrotemporal speech features by deriving statistical models of neural speech encoding from brain data measured with complementary meth-ods (fMRI and electrocorticography). Analysis of the derived model parameters will help to elucidate the neu-ral audiovisual integration mechanisms in human speech coding.
Dr. Christian Kluge ()
Prof. Dr. Hans-Jochen Heize (
Based on findings in animals that the auditory cortex has key capabilities to undergo plastic changes in response to learning and that the cholinergic system plays a key role in these processes, this project investi-gates the influence of cholinergic modulation on auditory cortical processing. Making use of the opportunity to directly influence the endogenous cholinergic system in a cohort of patients implanted with deep brain stimulation systems, and complemented by a pharmacological strategy in healthy control participants, we aim to further our understanding of the adaptive active sensing properties of the auditory system and their behavioral consequences.
Dept. of Neurology, University of Magdeburg
Dr. Thomas Brand ()
Dr. Stephan Ewert ()
Prof. Dr. Dr. Birger Kollmeier ()
This project aims at developing a comprehensive auditory model of human speech recognition. The challenge is to mimic human speech cue selection and processing based on mixtures of desired speech and interfering signals. The auditory model will be extended to use psychoacoustically and physiologically relevant cues such as envelope modulations, temporal fine structure and neural spike processing. Furthermore, methods of computational scene analysis (e.g., binaural source localization) and automatic speech recognition will be applied.
Dr. Volker Hohmann
For the localization, identification and separation of sound objects, the human auditory system binds in a largely unknown way several monaural and binaural signal-related feature streams to form a consistent image of the acoustical environment. The aim of the project is to clarify and model this Auditory Scene Analysis (ASA) process by combining statistical source encoding methods with psychoacoustically and physiologically motivated signal pre-processing. Binaural hearing in real listening environments including speech, noise and reverberation is particularly considered. Possible applications of the models are signal processing in hearing aids and automatic speech recognition.
Signal Processing for Hearing Aids, Medical Physics, Institute of Physics, University of Oldenburg
Prof. Dr. Jesko Verhey
The aim of the project is to understand the mechanisms underlying the perception of spectro-temporal characteristics of a sound within one ear and when using both ears. In the first part of the project, it is investigated what the minimal requirements for a stimulus are to detect and integrate coherent envelope fluctuations in different frequency regions. Then the across-frequency aspects of binaural processing are characterised and compared to the corresponding across-frequency processes within one ear. Finally, it is investigated how suprathreshold perception changes in conditions of masking release and how these conditions are represented in the EEG.
Prof. Dr. Hans Colonius
The correct estimation of the temporal order of auditory and visual stimuli and their duration in the sub-second range is a critical part of audiovisual object formation. Currently, there is no theoretical framework for the processing of time in a multisensory context required to understand crossmodal phenomena like “temporal ventriloquism” or the shift of the “time window of audiovisual synchronicity”. Based on applications of the theory of Fechnerian Scaling developed in the previous funding periods, this project focuses on developing and probing a modified paradigm of asynchronicity and temporal order judgment that promises to become the cornerstone of such a framework.
Prof. Dr. Dr. Birger Kollmeier ()
Dr. Bernd Meyer ()
The successful application of physiological and psycho-acoustic findings from several subprojects to automatic speech recognition (ASR) in B05 will be extended to design a context-aware ASR system. Both bottom-up and top-down features will be implemented based on computational auditory scene analysis (CASA) to operate in a complex acoustic scene. This framework will be used to predict sentence intelligibility of human listeners, to compare speech recognition of humans and machines, and to apply our findings in ASR system for unsupervised speech intelligibility tests.
Dr. Jörn Anemüller (
Dr. Stephan Ewert ()
Blind source separation (BSS) is a technical approach to acoustic multi-source/multi-microphone scene analysis that will be enhanced in this project using knowledge from psychoacoustics and physiology: BSS is used here as an active and adaptive auditory process that decomposes an auditory scene into its objects. In the first approach (WP 1), the output of an auditory model will be used to optimise parameters of a “technical” spatial filter. In the second approach (WP 2), object separation is performed on the output of an auditory model with processing kept as unconstrained as possible, leaving the extraction of object-binding features to the BSS, based on the statistics of the input stimuli. The goal is an improvement of BSS for technical applications as well as a better understanding of auditory object binding and segregation.
Prof. Dr. Steven van de Par
The main goal of this project is to investigate the role of top-down and bottom-up processing in binaural cocktail party settings. Bottom-up processing of binaural cues relates to the well-known binaural release from masking. Top-down processing relates to the contribution of binaural localization cues to auditory stream segregation which allows for selectively attending to one target only. Up till now there is little research that directly compares the contributions of these two aspects to binaural cocktail party processing. This project will use a new stimulus paradigm to get a better understanding of the contribution of both cues to the cocktail-party effect.