Signal Processing Algorithm Design Edge AI
+12 yrs Audio DSP 4 Patents 4 Awards
I'm an Audio Engineer specialising in Digital Signal Processing. For over a decade I've built algorithms that analyse, process and enhance speech, so voices sound clearer, more natural and easier to understand.
Since 2019, I've been part of Analog Devices, Inc. (ADI) in Leuven (Belgium), where I develop advanced voice processing algorithms for some of the world's most popular True Wireless Stereo (TWS) Earbuds, Headsets, Virtual Reality (VR) Devices and Smartwatches. My work has also led to the co-invention of 4 U.S. Patents on Audio DSP Systems and Methods.
Before joining ADI, I spent 2 years as an Audio DSP Researcher at Cirrus Logic in Melbourne (Australia), working on novel speech enhancement techniques using inertial measurement units (IMUs) and improving voice biometrics for widely used headsets and smartphones.
I hold a Bachelor's in Mechatronics and a Master's in Electrical & Electronic Engineering. My postgraduate research focused on extracting speech acoustic features to identify early indicators of depression in adolescents.
Today, my interests lie at the intersection of audio engineering, speech enhancement, machine learning, and human-centered technology — exploring how signal processing can make the way we communicate and interact with the world sound better.
“The best audio technology is often the one you never notice — until it stops working.”
Design and optimization of DSP algorithms operating in the frequency domain for efficient spectral processing and real-time audio analysis.
Algorithms and systems for improving speech quality and intelligibility in adverse acoustic conditions.
Beamforming and spatial processing for multi-microphone arrays in voice communication systems, enabling clear hands-free conversations.
Adaptive filtering and spectral methods for echo suppression in hands-free communication systems and conferencing applications.
Dynamic range control and spectral shaping for audio signals, from compression and expansion to limiting and equalization.
Audio processing for hearing assistance and enhancement applications.
Audio plugin development using the JUCE framework for professional audio applications, effects, and instruments.
Automated quality assessment metrics for speech processing algorithms, enabling fast and reliable algorithm evaluation.
Restores natural voice balance from an in-ear sensor when earbud fit or ANC distorts its spectrum, using a spectral-centre measure to drive a correction filter.
Mixes bone- and air-conducted voice through an adaptive crossover found where their cumulated spectra intersect, favouring the cleaner sensor without noise estimation.
Suppresses echo in earbuds, including non-linear components that traditional cancellers miss, with an MMSE-LSA estimator applied in the frequency domain.
Fuses bone-conduction, feedback and feedforward sensors through two adaptive crossovers that track ANC mode, echo and ambient noise.
Identifies acoustic markers of adolescent depression, including flatter intonation, reduced pitch variation and slower speech, as a basis for audio-based detection. Presented in Cairns, Australia.
Analog Devices
In recognition for outstanding effort and performance in tuning PureVoice on ADAU1860 + Maxims Buds Reference Design. This demo was successfully demonstrated at the ADI Hearables lounge bar at CES 2023 (Las Vegas).
RMIT University
$10k grant award for conducting research in the areas of emotion recognition from speech.
RMIT University
Sustained an average grade of 90% over 2 semesters for my postgraduate research.
RMIT University
Invited to present my thesis following successful completion of my postgraduate degree as a high distinction student.
RMIT University
Achieved a perfect score of 100% in project planning, preparation and problem-solving.
RMIT University
“Smart Mining Helmet” selected amongst 63 teams for best innovative design. The EnGenius trade fair highlights a diverse range of innovative engineering solutions and provides students with the opportunity to network with invited industry representatives. Student projects are assessed by a panel of industry sponsors, with prizes awarded for the best innovative designs.
Golden Key International Honour Society
Recognised as being amongst the top 15% of postgraduate students. Golden Key has been working with RMIT University since 1993 to recognise and reward high-achieving students.