Keynote speakers:

Topic and Bio

Dr. Dorothea Kolossa

Professor, Electronic Systems of Medical Engineering
TU Berlin

Audio-visual technologies for inclusion: Bringing speech intelligibility beyond the limits of hearing

Hearing impairments can profoundly affect speech communication, often leading to social isolation, emotional distress, and cognitive decline, despite advancements in hearing aids and cochlear implants. In challenging environments where audio-only communication reaches its limits, multimodal approaches that integrate video with audio offer promising solutions. Recent innovations in audio-visual speech technology are shifting the field beyond traditional auditory methods, enabling better speech comprehension even in difficult listening conditions. This presentation will explore these advancements in audio-visual speech enhancement, video-based support for hearing, and modeling the perceptual basis of audio-visual speech intelligibility. It will conclude with a brief look at future research directions, including the key questions of privacy-preservation and responsible innovation.

Dorothea leads a research group at TU Berlin that, among other topics, is working on applications that use audiovisual input for highly robust speech recognition and video-based speech quality enhancement. The group uses adversarial examples to examine the robustness of deep-learning-based speech recognition.

Dr. Giovanni Di Liberto

Assistant Professor in Intelligent Systems
School of Computer Science and Statistics
Trinity College Dublin, the University of Dublin

Investigating auditory cognition with natural speech and music

Perception involves making sense of the world around us by processing a continuous flow of sensory information. In doing so, the human brain produces electrical activity that synchronises to particular properties of sensory inputs, a phenomenon referred to as neural tracking. The case of auditory perception is particularly remarkable. The discovery that neural signals reliably track the amplitude envelope of continuous sounds has led to new research directions, such as the study of auditory attention in immersive multi-talker scenarios. Here, I will present a series of studies investigating the neural tracking phenomenon when listening to speech and music, wherein both low-level acoustic properties and more abstract linguistic structures have to be processed. I will demonstrate methodologies for disentangling neural responses to stimulus properties at different abstraction levels from a single electroencephalography (EEG) recording. I will then describe recent developments where the resulting objective indices are used for probing auditory perception across various contexts, such as language learning and developmental research.

Giovanni is an Assistant Professor in Intelligent Systems the School of Computer Science and Statistics, Trinity College Dublin since 2021. He is also a PI at the Trinity College Institute of Neuroscience (TCIN) and the ADAPT Centre. Giovanni’s research focusses on understanding  the brain mechanisms underlying speech comprehension and music processing, using computational methods grounded in computer science and engineering. Giovanni leads a multidisciplinary research team, with backgrounds including Computer Science, Engineering, Cognitive science, Psychology, and Linguistics.

Dr. Brent Edwards

Director of the National Acoustic Laboratories
(NAL)
Sydney, Australia

Revolutionizing Audiological Care with AI

Artificial Intelligence (AI) is transforming every industry that exists. This rapid development of AI innovation is changing all aspects of healthcare and will have a dramatic impact on hearing healthcare services. Hearing aids AI chips are now becoming available with dramatic capabilities to benefit people with hearing difficulty. AI tools are being produced to help with access and effectiveness of health care. All these changes present both challenges and opportunities for hearing care providers for integrating AI into their practice. This presentation will provide an overview of the application of AI in healthcare over the past several years and what its implication is for hearing healthcare: which aspects of hearing healthcare AI has the opportunity to transform, how the role of audiology will evolve as AI is integrated in hearing services, and ethical considerations in the use of AI in hearing healthcare. Research at NAL in these areas will be used to provide insight into understanding these changes and provide guidance on how hearing care professionals can manage this evolution.

For over 22 years he headed research at major hearing companies (GN ReSound, Starkey) and at Silicon Valley startups that have developed innovative technologies and clinical tools used worldwide. Dr. Edwards founded and ran the Starkey Hearing Research Center in Berkeley, California that was a leading site for research in hearing impairment and cognition. He is a Fellow of the Acoustical Society of America, a Fellow of the International Collegium of Rehabilitative Audiology and an Adjunct Professor at Macquarie University.

Dr. Louise Hickson
Professor of Audiology and Executive Dean
of the Faculty of Health and Behavioural
Sciences at The University of Queensland

Exploring the social connections of adults with acquired hearing loss.

Communication underpins our social connections, supporting our participation and engagement in society. Acquired hearing loss can impact communication, and subsequently, social connection, resulting in social isolation, loneliness, and poor wellbeing. In this presentation I will present a snapshot of a current study that is exploring the social connections of adults who present to audiology clinics for the first time about their hearing loss.  Data is collected online using traditional surveys and real world video recorded conversations, as well as a new method in the field – online Social Identity Mapping (oSim).   We are collecting data at baseline, 3-months, 6-months, and 12-months and we aim to understand the factors, including social connection, that influence hearing aid uptake for adults with hearing loss.

Professor Hickson has published over 290 research articles, books and book chapters with her main focus on the effects of hearing loss on people’s everyday lives and the development of strategies and interventions that improve the uptake and outcomes of hearing rehabilitation. Professor Hickson is Past President of the International Society or Audiology and a Fellow and Past President of Audiology Australia. Continue reading.

Featured talks:

Topic and Bio

Fotios Drakopoulos, PhD

Research Fellow, Ear Institute
University College London

Computational models of the cochlea are highly accurate and widely used across academia and industry, but models of the auditory brain lag far behind in both performance and penetration. Here, I will demonstrate how recent advances in large-scale electrophysiology and deep learning can be combined to develop accurate models of sound coding in the brain. I will present the recently developed ICNet model, which provides near-perfect simulation of neural responses in the gerbil inferior colliculus (IC). ICNet takes a sound waveform as input to produce activity patterns across thousands of neural units with millisecond precision, capturing more than 86% of the explainable variance in neural activity across a wide range of sounds. The proposed ICNet model is fast to execute and provides a compact representation of central auditory processing through its latent dynamics, thus facilitating a wide range of hearing and audio applications.

Fotios received his Ph.D. in electrical engineering from Ghent University, as part of the Hearing Technology Lab led by Prof. Sarah Verhulst. Since 2023, he is a Research Fellow at the UCL Ear Institute under the supervision of Prof. Nicholas Lesica. His research has been focussed on the application of advanced computational techniques to improve real-world applications such as hearing aids. His main research interests include auditory perception, machine hearing, audio signal processing, computational auditory modelling and psychoacoustics.

Malcolm Slaney, PhD

Stanford University

Malcolm will provide an overview of emerging machine learning and AI techniques in auditory modeling, comparing domain-specific auditory models with large-scale data-driven approaches. Drawing from decades of work at the intersection of auditory perception, signal processing, and machine learning, Malcolm will illustrate how current techniques are reshaping our ability to simulate and understand the human hearing system.

Dr. Malcolm Slaney is a Consulting Professor at Stanford and Affiliate Faculty at the University of Washington, known for his work in auditory perception, signal processing, and machine learning. He co-authored Principles of Computerized Tomographic Imaging and created the widely used Auditory Toolbox in MATLAB. He has held research roles at Google, Microsoft, Yahoo!, and Apple, contributing to audio analysis and multimedia search. An IEEE Fellow and recipient of the 2021 Misha Mahowald Prize, he is active in the Telluride Workshop and currently explores brain-audio interfaces.

Marina Salorio Corbetto, PhD

University of Cambridge

The BEARS project is a national UK clinical trial evaluating a virtual-reality training suite for children and young people who use bilateral cochlear implants. It is the first NIHR programme looking at a behavioural intervention for deaf children. The objective of the project is to assess the efficacy of the BEARS training suite to improve speech in noise and sound localisation outcomes. Additionally, new clinical tools for assessments of spatial listening skills have been developed within the project, making it possible to implement complex hearing assessments with simple set-up and space requirements.

Marina Salorio Corbetto is a research associate at SOUND Lab, Prof. Deborah Vickers’ laboratory at the Department of Clinical Neurosciences of the University of Cambridge. She is an audiologist who has worked with people with hearing aids and cochlear implants. Her work  focuses on improving outcomes for these groups.

Dietmar M. Wohlbauer, Ph.D.

Harvard Medical School and Massachusetts Eye and Ear. Department of Otolaryngology, Head and Neck Surgery. Boston, MA, USA.

Neural sensitivity in cochlear implant stimulation refers to the activity of auditory neurons following electric stimulation. We investigated how physiological recordings and perceptual measures interact and how their relationship might reflect the quality of the electrode-neuron interface (ENI) as a possible indicator for the neural survival along the implanted cochlea.

Dietmar received his Ph.D. in neuroscience from the University of Zürich, Switzerland, in 2021. He is currently a Postdoctoral Fellow at the Harvard Medical School and the Massachusetts Eye and Ear in Boston, MA, USA. His research interests focus on unilateral and bilateral cochlear implant sound coding, spatial perception with hearing prostheses, and investigating the associated underlying mechanisms via physiological and perceptual assessment methods.

Caleb Rascon, Ph.D.

Instituto de Investigaciones en
Matemáticas Aplicadas y en Sistemas
Universidad Nacional Autónoma de México

Potential Impact of Real-time Auditory Scene Analysis on Hearing Aid Research

Auditory Scene Analysis aims to localize, separate, and classify sound sources in a given acoustic environment, effectively trying to emulate close-to-human hearing on a non-human entity. This can highly benefit Hearing Aid Research, however, there are important hurdles and challenges in doing so, such as the need to carry out the aforementioned tasks in real-time, in a highly dynamic environment, but only using minimal computational resources. In this talk, a novel multi-agent approach is proposed to address these challenges, as well as several tasks that could be carried out to further enhance its viability in Hearing Aid Research.

Sheila Templado, Audiologist

Templado Advanced Audiology
and Otolaryngology Clinic. Murcia, Spain.

Frequency-Following Responses and Late Latency Responses as Neurobiological Markers in Fetal Alcohol Spectrum Disorders