Curriculum vitae
Things I have been doing
Work History
Google, Deepmind
July 2024 - Present
Research Engineer — Neural audio processing.
LY Corporation (LINEヤフー株式会社), Music Processing Team
September 2023 - June 2024
Researcher — Research and development of speech processing algorithms.
LINE Corporation, Speech Team
March 2020 - September 2023
Senior Researcher — Research and development of speech processing algorithms.
SAFECAST Japan
March 2011 - around 2020
bGeigie dude — Volunteer, designing, manufacturing and writing firmware for car-borne geo-tagged radiation measurement devices as part of a grass-root initiative to monitor fallout from the Fukushima Dai-ichi triple meltdown of March 2011. So far, 15 million measurement points have been collected using this system.
Tokyo Metropolitan University, Ono Laboratory
October 2019 - February 2020
Project Associate Professor — Sound-to-light conversion sensors (Blinkies), and blind source separation with many microphones.
Tokyo Metropolitan University, Ono Laboratory
October 2017 - September 2019
Postdoctoral Fellow — We use sound-to-light conversion sensors to easily build large sound sensing arrays. Applications are in conventional audio processing and large scale monitoring.
EPFL, Audiovisual Communications Laboratory
April 2017 - September 2017
Postdoctoral Researcher — Supervision of students. Prospecting data providers and participants for the Data Jam Days 2017 event at EPFL. Research and experiments in array audio processing.
EPFL, Audiovisual Communications Laboratory
September 2012 - March 2017
Research Assistant — Investigating how structural aspects of sound propagation, the echoes in a room for example, can provide more robust and efficient algorithms for problems such as beamforming, direction of arrival, or source separation. In a different line of work, I apply techniques from information theory and theoretical computer science to provide low-complexity algorithms for signal processing problems.
NEC Corporation, Media & Information Processing Laboratory
January 2011 - August 2012
Research Engineer — Development of new algorithms for single-microphone speech denoising techniques based on spectral subtraction for application in consumer electronics. Specifically, speech harmonics preservation, DCT-based spectral subtraction, wind noise suppression.
IBM Research — Zürich, Systems Software Group
February 2009 - September 2010
Research Engineer — Developed algorithms for the computation of the continuous Haar, Fourier and cosine transforms of rectilinear polygons found in IC layouts. Derived the theoretical computational complexity. Implemented the algorithms in a lithography tool and benchmarked against discrete transforms. Investigated the sparsity of the representation of the polygons in Haar basis and its application to the pattern analysis of IC layouts. Outcome resulted in scientific articles, highly optimized software and patents filing.
NEC Corporation, System IP Core Laboratory
March 2007 - January 2008
Research Intern — Implementation in Matlab of a 3GPP compliant simulator to evaluate the performance of different interference mitigation algorithms. Implementation and performance evaluation of QRD-RLS interference mitigation for Wideband CDMA on a CORDIC-based reconfigurable systolic array for software-defined radio (SDR) applications. The result of this work was presented at a domestic SDR conference in Nagano, Japan.
Education
Ph.D. in Signal Processing
EPFL, Lausanne, Switzerland, 2017
Thesis: Rake, Peel, Sketch - The Signal Processing Pipeline Revisited pdf
Thesis: Rake, Peel, Sketch - The Signal Processing Pipeline Revisited pdf
M.Sc. in Communications Systems
EPFL, Lausanne, Switzerland, 2009
Thesis: Algorithms for the Computation of Continuous Transforms of Rectilinear Polygons pdf
Thesis: Algorithms for the Computation of Continuous Transforms of Rectilinear Polygons pdf
B.Sc. in Communications Systems
EPFL, Lausanne, Switzerland, 2006
Erasmus exchange
Royal Institute of Technology, Stockholm, Sweden, 2006
Skills
Languages
- French (Native)
- English (Fluent)
- Japanese (Fluent)
- German (Conversational)
- Swedish (Basic)
Expertise
-
Signal processing,
Neural networks,
Multi-channel audio processing,
Blind source separation,
Optimization,
Algorithms,
Fourier,
Embedded systems
Programming
-
Python,
C,
C++,
Latex,
git
Operating Systems
-
Unix,
Mac OS X,
Windows
