Currently Funded Research Projects

The MARCS Institute for Brain, Behaviour and Development has a number of funded research projects running at any given point in time.

All MARCS research projects aim to have a profound impact in their particular field and often involve collaboration with local and international researchers.

Below is a detailed list of each currently funded project, complete with a relevant problem and solution that gives a brief overview of the research.

Biomedical Engineering and Neuroscience Program (BENS)

 
1.    Hardware implementation of spiking neural network architectures for sensory and learning systems


Problem: Current Micro-Aerial Vehicles have inadequate agility, manoeuvrability, and navigational capabilities (sensors and algorithms) for safe, effective operation in cluttered urban environments.

Solution: We will develop novel computer vision algorithms for neuromorphic event-based sensors to create a highly efficient, activity-driven system that can meet the power-, weight-, and size-constraints of MAVs.

 Prof. André van Schaik, Saeed Afshar
Department of Defence 2016 – 2019 


 2.    Hardware Acceleration for Neural Systems
 
Problem: To really understand how brains work, we need to simulate neural networks of a size larger than possible on standard computers. Several expensive electronic systems are being developed world wide to allow this but the output from these projects will only be accessible to a few researchers at the Universities where this work is being done.

Solution: Our researchers are developing such an electronic system using cheaply available commercial hardware and they will make their design and software available for free, so that many brain researchers worldwide may benefit from this capability to enhance their understanding of the brain.

 Prof. André van Schaik, Dr. Mark Wang
ARC Discovery Project 2014 – 2016


3.    Event-based image sensors for optical space situational awareness

Problem: DST Group is developing a ground based Space Situational Awareness capability to monitor objects in space above Australia. Current CCD image sensors have a low dynamic range, and their long exposure time makes them unable to image fast moving objects.

Solution: Event-based cameras have a much higher dynamic range and temporal resolution than CCD cameras, and we are developing novel image processing algorithms to enable these event-based cameras to be used for space situational awareness applications.

 Prof. André van Schaik, Dr. Gregory Cohen
Department of Defence 2016 - 2017 


 4.    Event Based Sensor Space Situational Awareness Experiment
 
Problem: DST Group is developing a ground based Space Situational Awareness capability to monitor objects in space above Australia. Current CCD image sensors have a low dynamic range, and their long exposure time makes them unable to image fast moving objects.

Solution: Event-based cameras have a much higher dynamic range and temporal resolution than CCD cameras, and we are developing novel image processing algorithms to enable these event-based cameras to be used for space situational awareness applications.

  Prof. André van Schaik, Dr. Gregory Cohen
 Department of Defence 2017 


 5.    Design of a space-based, low-power star tracking system using event-based cameras 


Problem:
Current computer vision systems do not have the dynamic range to deal with the range of brightness of objects encountered in space, or the temporal resolution to deal with the high speed of some of these objects. Standard cameras also have a very high data rate, which is incompatible with the limited communication bandwidth of a small satellite.

Solution: Event-based cameras have a much higher dynamic range and temporal resolution, and much lower data rates than standard cameras, and we are developing novel image processing algorithms to enable these event-based cameras to be used for space situational awareness applications.

 Prof. André van Schaik, Dr. Gregory Cohen
Department of Defence 2017 - 2018 


6.    Neuromorphic Cognition for Micro-Aircraft Applications

Problem: Current Micro-Aerial Vehicles have inadequate agility, manoeuvrability, and navigational capabilities (sensors and algorithms) for safe, effective operation in cluttered urban environments.

Solution: We will develop novel computer vision algorithms for neuromorphic event-based sensors to create a highly efficient, activity-driven system that can meet the power-, weight-, and size-constraints of MAVs.

 Prof. André van Schaik, Dr Gregory Cohen
Department of Defence 2016 – 2017 


 7. ARC Training Centre in Cognitive Computing for Medical Technologies [via University Melbourne]

Problem: There is a lack of a trained workforce in Australia and abroad with expertise in using cognitive computing technologies in medical contexts.

Solution: This project aims to create a workforce that is expert in developing, applying and interrogating cognitive computing technologies in data-intensive medical contexts, to facilitate the next generation of data-driven and machine learning-based medical technologies. The Centre will provide a world-class industry-driven research training environment for PhD students and postdoctoral researchers, who will lead the medical technology industry into a new era of data-driven personalised and precision medical devices and applications, based on a combination of research, industry and entrepreneurial skills that they will acquire through the Centre.

  Prof. André van Schaik
 ARCG Industrial Transformation Training Centres 2017 - 2022  


 8.    Clinical Trials of the First Bionic Voice Prosthesis for Voice-loss Children

Problem: Despite the emerging progress in many fields of bionics and bio-signal processing, laryngectomy patients still rely on the gold standard of the TE voice prosthesis which is intrinsically limited to let them step beyond a hoarse, whispered voice.

Solution: This project designs the first functional Bionic Voice prosthesis for children with Tracheostomy. The proposed Bionic Voice is a non-invasive electronic solution, which substitutes the voice generation function of an amputated or bypassed larynx.

 Dr. Farzaneh Ahmadi, Alan Cheng, Daniel Novakovic
The Children's Hospital at Westmead 2016 - 2017  


 9.    Designing a Bionic Voice Prosthesis to Restore Natural Voice for Laryngectomy Patients

Problem: Despite the emerging progress in many fields of bionics and bio-signal processing, laryngectomy patients still rely on the gold standard of the TE voice prosthesis which is intrinsically limited to let them step beyond a hoarse, whispered voice.

Solution: We will design the first bionic voice prosthesis controlled by respiratory system and the neural/neuromuscular drive of the missing larynx using novel signal processing algorithms and a miniature loudspeaker to provide a more naturally sounding voice.

 Prof. André van Schaik, Dr. Farzaneh Ahmadi
The Garnett Passé and Rodney Williams Memorial Foundation Fellowship 2015 - 2019  


 10.    Use of a Portable Stimulator to Treat GWI [via Veterans Bio-Medical Research Institute]
 
Problem: Gulf War Illness is a multi-symptom illness with nausea, headache and dizziness among them.
 
Solution: Impaired inner ear function is a likely cause of these symptoms and a patented stimulation technique we have developed could provide relief to the sufferers.

 Prof. Jorge Serrador, Dr. Paul Breen, Michael Schubert, Scott Wood, Prof. Gearóid Ó Laighlin
 US Department of Defense (CDMRP) 2014 - 2016  


11.    Treatment of Vestibular Dysfunction using a Portable Stimulator

Problem: Veteran soldiers have trouble maintaining balance and performing complex functions such as driving due to traumatic brain injury.

Solution: Restore this function by stimulating the balance organs using a novel neural enhancement technique developed by researchers in the MARCS Institute.

 Prof. Jorge Serrador, Dr. Paul Breen, Michael Schubert, Scott Wood, Prof. Gearóid Ó Laighlin
US Department of Defense (DMRDP) 2014 – 2018  


 12.    Enhanced Sensory Perception via Jitter Reduction and Neural Synchronisation Evoked by Subsensory Electrical Noise Stimulation - Restoring Sensitivity in Peripheral Neuropathy

Problem: There are over 100 different forms of peripheral neuropathy and very limited treatment options. In the elderly, reduced peripheral sensation is associated with recurrent falls and fractures. In diabetes, peripheral neuropathy is a significant risk factor for falls, foot ulceration and ultimately amputation. Restoring sensory perception would have a major impact on the quality of life and prognosis for those with peripheral neuropathy.

Solution: Using a sub sensory form of electrical stimulation we can modulate the remaining somatosensory nerve activity and enhance sensory perception, an innovation that is unique to this research team. We have already shown in computational models and in humans, that subsensory electrical noise stimulation (SENS) can improve neural function. Using this intervention we can consistently enhance sensory function by as much as 16% in young adults, older adults and patients with diabetes. Further enhancements in complex functions such as proprioception, balance and gait have also been recorded.

 Dr. Paul Breen, Dr. Klaus Stiefel, Dr. Ingvars Birznieks, Dr. Stefania Penkala
NHMRC Project 2014 - 2017  


 13.    Microelectronic 3D Imaging and Neuromorphic Recognition for Autonomous UAVs [via Milan Politechnic]

Problem: There is a need to provide vision in low-light conditions to unmanned aerial vehicles (UAVs). Miniaturisation and power constraints for sensing using an array of Single Photon Avalanche Diodes (SPAD) on UAVs are challenging.

Solution: We will develop a neuromorphic algorithm for view-invariant 3D image recognition, and its hardware implementation to be used with the SPAD sensor array.

 Dr. Tara Hamilton, Dennis Delic, Jean-Michel Redoute, Franco Zappa
NATO Grant Science for Peace and Security Programme 2016 - 2019  


 14.    Microelectronic 3D Imaging and Neuromorphic Recognition for Autonomous UAVs
 
Problem: There is a need to provide vision in low-light conditions to unmanned aerial vehicles (UAVs). Miniaturisation and power constraints for sensing using an array of Single Photon Avalanche Diodes (SPAD) on UAVs are challenging.

Solution: We will develop a neuromorphic algorithm for view-invariant 3D image recognition, and its hardware implementation to be used with the SPAD sensor array.

 Prof. André van Schaik, Dr. Dennis Delic, Dr. Mark Wang
Department of Defence 2016 - 2019



Speech and Language Learning

1.    ARC Centre of Excellence for the Dynamics of Language

Problem: What we know about language – how it is patterned, learned, processed and adapted – is disproportionately influenced by research into major languages like English in monolingual, Western cultural settings, yet Australia sits at the epicentre of the world's linguistic diversity.

Solution: The ARC Centre of Excellence for the Dynamics of Language, led by ANU with partners UWS, UQ and UniMelb, will transform our understanding of the languages of the Indo-Pacific region and address challenges raised by language diversity for education, social access and well-being, and support the development of communication and learning technologies vital for a truly multilingual and future-proofed Australia.

 Prof. Anne Cutler, A/Prof. Paola Escudero, A/Prof. Caroline Jones
ARC Centre of Excellence for the Dynamics of Language 2014 - 2020


2.    Cracking the code of successful language learning

Problem: It is not clear why some second language learners flourish, while many struggle in the same educational setting.

Solution: Identify the combination of factors that matter most in successful language learning, with the aim of tailoring training proactively to maximise learning outcomes.

 Dr. Mark Antoniou
ARC DECRA Fellowship 2015 - 2017


3.    Explaining the native-language listening advantage by charting neural response and perceptual adaptation across languages - but within individuals

Problem: Why do we listen to speech better in our native language?

Solution: This project investigates i) whether listeners better adjust to specific talkers' speech sounds, and more efficiently adjust how they recognise words, in their native than in their second language, and ii) how the brain might give native language its listening advantage.

 Prof. Anne Cutler
ARC Discovery Project 2014 - 2016


 4.    The seeds of literacy in infancy: empirical specification of the acoustic determinants of language acquisition

Problem: Currently dyslexia cannot be diagnosed until reading instruction begins.

Solution: Devise diagnostic tests appropriate for 6-month-old infants, administer these to infants not-at-risk and at-risk for dyslexia (due to having a dyslexic parent), and follow them through until reading age.

 Prof. Denis Burnham, Prof. Usha Goswami
ARC Discovery Project 2011 - 2015


5.    Early precursors of linguistic development

Problem: Delayed or deficient linguistic abilities can be particularly debilitating, but such problems are not evident until children are two to three years of age, and in the case of written language, until they are five or six years.

Solution: Conduct longitudinal studies from early infancy (six months) through to onset of school across different language backgrounds investigating brain functions, auditory skills, parental speech input, and how the infant brain reacts to hearing aids and cochlear implant input in order to identify critical developmental factors & plan for intervention procedures.

 Prof. Denis Burnham
ARC Extension to the HEARing CRC 2014 – 2019


6.    Alveo: Human communication science virtual laboratory

Problem: It is difficult for a researcher from one discipline to apply the tools and techniques of another discipline, or to explore data collected under one paradigm via a completely different analytical perspective. Moreover, research conducted in isolation entails inefficient repetition of analysis of local data sets.

Solution: The Alveo Virtual Laboratory provides easy access to a variety of databases and the integrated use of a range of analysis tools, from linguistics, natural language processing, speech science, psychology, music and acoustic processing. It will foster inter-disciplinary research and facilitate the discovery of new methods for solving old problems.

 Prof. Denis Burnham, Dr. Dominique Estival
NeCTAR Grant 2013 – 2019


7.    Understanding bilingual language acquisition in northern Indigenous Australia: phonological, lexical, orthographic, and family factors

Problem: How do young Indigenous children in northern Australia grow up bilingual in English and in Kriol, an English-based creole language?

Solution: This research will establish the linguistic, educational, and family factors in successful language acquisition for these children, with implications for long-term health, social inclusion, education and employment.

 A/Prof. Caroline Jones
ARC Future Fellowship 2012 – 2017


 8.    Understanding different speakers vs. different accents: apples and apples or pears

Problem: Speech perception is possible because humans are able to normalize the variability in the speech signal. To date, we do not know whether the underlying skills to cope with different speakers and accents are uniquely human and specific to certain languages.

Solution: Researchers at MARCS will use innovative technology to unveil the details of an important component of human cognition, namely speech comprehension. The findings will lay foundations to advance both automatic speech recognition systems and hearing aids for human communication.

 A/Prof. Paola Escudero
ARC Discovery Project 2013 – 2015


9.    You came TO DIE?! Perceptual adaptation to regional accents as a new lens on the puzzle of spoken word recognition

Problem: How do we achieve stable word recognition across the pronunciation differences of unfamiliar accents?

Solution: Studies will offer a crucial insight into how listeners relate variations in a word's spoken form to its underlying constant identity. Sensitive perceptual tests and computational modelling of cross-accent word recognition by speakers of Australian, New Zealand, and UK English will provide new insights for theories of spoken word recognition, and improve understanding of why unfamiliar accents pose problems for automatic speech recognisers, second language learners and listeners with language impairments and comprehension limitations.

 Prof. Catherine T. Best, Dr. Jason Shaw, Prof. Gerry Docherty, Prof. Jen Hay, Prof. Paul Foulkes, A/Prof. Bronwen Evans
ARC Discovery Project 2012 – 2014


10.    Children's generalisation and adaptation to unfamiliar regional accents reveal the path of early word learning

Problem: Young children recognise that unfamiliar people may speak with a different accent than they hear in their homes. What is not yet understood is how they learn to recognise words spoken in unfamiliar accents.

Solution: Our studies of infants' and toddlers' early word-learning will track the emergence of flexible word recognition, from early dependence on the accent they hear in their homes, to gradual improvement in generalisation to words spoken in unfamiliar accents

 Prof. Catherine T. Best, Dr. Christine Kitamura
Marcus and Amalia Wallenberg Foundation 2013 – 2015


11. Enhancing language learning via auditory training and interaction

Problem: Learning to understand a foreign language is difficult for adults and children. Foreign sounds that distinguish words and that are not found in the learner's native language are particularly difficult, yet they are the building blocks for communication.

Solution: Infant-parent interaction is likely to help language learning for both infant and adult, as it aids cognitive and linguistic development. The project will capitalize on the benefits of infant-parent interaction when presenting a language learning tool with targeted stories and auditory training that aims at enabling and enhancing foreign language processing in both parents and children.

 A/Prof. Paola Escudero
ARC Future Fellowship 2017 – 2020


 12.    Come on kids, pay attention to your prosody

Problem: A conversation between two speakers seems effortless, yet it requires processing of the incoming speech stream while simultaneously preparing a response and anticipating the exact moment it is appropriate to talk. Adults' knowledge of their native language and the communicative process allows them to complete this amazing feat. To date, it is unknown how children achieve this challenging task given that during the first years of their live, their native language competence is still developing.

Solution: This is the first study to assess the skill of conversational turn-taking in the first year of life, and across development, to not only understand the emergence of this ability but also how it integrates with children's growing linguistic knowledge. Twelve-, 24-, 36-month-old children's perception of turn taking is assessed experimentally via recordings of their anticipatory eye movements in an eye-tracking paradigm.

 Dr Marina Kalashnikova
Australian Linguistic Society 2017