Postgraduate Research in Neural Information Processing The Bionic Ear Institute is offering a number of Masters and PhD opportunities for students with honours degrees in the fields of Engineering, Physics, Mathematics, Psychology, Computer Science or related fields. The Bionic Ear Institute is an independent not-for-profit research organization affiliated with The University of Melbourne and therefore all degrees are offered through the University. Research at the Institute specializes in finding new and better ways to restore brain function as well as helping deaf children and adults to communicate. Research Activities The Bionic Ear Institute undertakes research to develop devices such as the Bionic Ear (cochlear implant) that enables people who are hearing impaired to hear by electrically stimulating the auditory nerve. New research directions at the Institute are extending this technology to develop other neural prostheses for spinal cord and nerve repair. Postgraduate research projects are available in a whole range of areas outlined below that involve the application of mathematical and engineering techniques to investigate how information is processed in biological neural systems (bio-inspired signal processing) and the development of artificial systems that mimic biological systems (biomimetics). Postgraduate projects are available in each of the areas below and are developed in close consultation with the postgraduate student, taking into account their interests, background and technical skills. Examples of possible projects in each of the areas are given.   Psychophysics, Bio-inspired Signal Processing and Biomimetics Computational Neuroscience and Neural Modelling: The Neural circuits in the brain are able to perform complex computations and learn about new aspects of the world. Furthermore they do so in a robust and reliable fashion.  Mathematical modelling can be used to investigate how these mechanisms transform sound into brain signals and ultimately lead to percepts such as sound location, pitch, timbre and speech recognition. Neural modelling research at the Institute is concerned with how neurons and neural systems (i) compute useful information, (ii) learn and adapt to their inputs and environment, and (iii) control and modulate their responses. Example of a PhD project in computational neuroscience and neural modelling. Psychophysics of the Auditory System: Psychophysics is the quantitative study of the relationship between sensations (psychological events) and the stimuli that produced them (physical events). Psychophysical study of the auditory system attempts to quantify how we perceive sounds and combinations of sounds, so that we can better understand the mechanisms of the ear and brain. Better understanding of how hearing works is used to aid the development of improved hearing aids and cochlear implants for helping people who have impaired hearing. Example of a PhD project in psychophysics of the auditory system. Cochlear Implant Speech Processing: The electrical stimulation of the auditory nerve by a cochlear implant aims to mimic the signal produced by sound (acoustical stimulation) in normal hearing. However hardware and software limitations in the presently available implant and the lack of an exact description of the signal produced acoustically results in many people with cochlear implants not hearing as well as normal listeners. Our group is seeking to close this gap by designing electrical stimulation strategies that will more closely mimic acoustical stimulation.  To do this we mathematically model the biophysical mechanisms that transform sound into brain signals and use this as the basis for developing speech processing algorithms. Example of a PhD project in cochlear implant speech processing. Adaptive Learning in Neural Architectures: The goal of this project is to develop an improved pattern recognition technique based upon networks of spiking neurons analogous to the brain. Learning takes place by timing-dependent changes to the strengths of the connections between neurons (synapses), as has been observed in recent physiological studies. The significance of this approach is that it directly incorporates fine-grained timing information that is not used by existing techniques. The expected outcome is an adaptive learning technique that recognises timing-dependent patterns, is robust to noise and operates in real-time. The technique will be applied to automatic speech recognition, where fine-grained temporal cues play an important role. Example of a PhD project in adaptive learning in neural architectures. Automatic Speech Recognition (ASR): Automatic speech recognition is a challenging problem involving the segmentation and classification of complex temporal and spectral patterns.  The presence of background noise makes the problem even harder.  Conventional ASR algorithms perform very poorly in noisy situations, such as when the talker is using a mobile phone.  Our group is seeking to develop robust ASR through the use of insights into how the human auditory system processes speech.  New models of the auditory pathway and spatiotemporal learning algorithms are applied to develop new ASR algorithms or improve existing systems. Example of a PhD project in automatic speech recognition. Further information Information about the academic entry requirements for postgraduate research degrees at The Bionic Ear Institute, as well as postgraduate scholarships, is available on the Postgraduate Information page. Additional information about scholarships and eligibility, as well as a description of some special scholarships, is available at the University of Melbourne postgraduate scholarships web page. More information about details of the degrees offered through the University of Melbourne may be obtained through the site for University of Melbourne Graduate Studies Program.  More information about the research undertaken at the Institute is available at the web site of The Bionic Ear Institute. Students who are interested in postgraduate research associated with any of the above projects should first get in touch with us, either by contacting the Postgraduate Coordinator (pgcoord@bionicear.org) or by emailing us directly: Dr Anthony N. Burkitt (aburkitt@bionicear.org) (background in Physics, Mathematics and Psychology) Computational Neuroscience, Neural Modelling, Adaptive Neural Learning, Cochlear Implant Speech Processing Dr David B. Grayden (dgrayden@bionicear.org) (background in Electrical Engineering and Computer Science) Automatic Speech Recognition, Cochlear Implant Speech Processing, Psychophysics