Bence Halpern

I’m a machine learning researcher focusing on speech technology for pathological speech. I have worked on a broad range of projects including automatic speech recognition, voice conversion, speech enhancement, acoustical signal processing and speaker recognition.

I’m currently a research engineer at Oxford Wave Research Ltd, a visiting researcher at Nagoya University.

I’m learning Japanese language as a hobby. 日本語土手じゃないです。

Journal papers

Low-resource automatic speech recognition and error analyses of oral cancer speech
Speech Communications
Bence Mark Halpern, Siyuan Feng, Rob van Son, Michiel van den Brekel, Odette Scharenborg
Paper

Conference proceedings

The Effectiveness of Time Stretching for Enhancing Dysarthric Speech for Improved Dysarthric Speech Recognition
Interspeech 2022
Luke Prananta, Bence Mark Halpern, Siyuan Feng, Odette Scharenborg
Paper

Mitigating bias against non-native accents
Interspeech 2022
Yuanyuan Zhang, Yixuan Zhang, Tanvina Patel, Bence Mark Halpern, Odette Scharenborg
Paper

Towards Identity Preserving Normal to Dysarthric Voice Conversion
ICASSP 2022
Wen-Chin Huang, Bence Mark Halpern*, Lester Phillip Violeta, Odette Scharenborg, Tomoki Toda
Paper

Pathological voice adaptation with autoencoder-based voice conversion
Speech Synthesis Workshop 2021
Marc Illa, Bence Mark Halpern*, Rob van Son, Laureano-Moro Velazquez, Odette Scharenborg
Paper

An Objective Evaluation Framework for Pathological Speech Synthesis
ITG Speech Communication
Bence Mark Halpern, Julian Fritsch, Enno Hermann, Rob van Son, Odette Scharenborg, Mathew.-Magimai Doss
Paper

Detecting and analysing spontaneous oral cancer speech in the wild
Interspeech 2019
Bence Mark Halpern, Rob van Son, Michiel van den Brekel, Odette Scharenborg
Paper

Residual networks for resisting noise: analysis of an embeddings-based spoofing countermeasure
Speaker Odyssey 2022
Bence Mark Halpern, Finnian Kelly, Anil Alexander
Paper

Conference abstracts

Speaker-informed speech separation and enhancement
IAFPA 2021
Bence Mark Halpern, Finnian Kelly, Anil Alexander
PPTX PDF

Can DeepFake voices steal high-profile identities?
IAFPA 2022
Bence Mark Halpern, Finnian Kelly
ABSTRACT POSTER

T. Tienkamp, T. Rebernik, B. Halpern, R. van Son, S. A. de Viscscher, M. Witjes, and M. Wieling
Speech Motor Control 2022
Quatifying changes in articulatory working space following oral cancer treatment using electromagnetic articulography
ABSTRACT: TBA POSTER: TBA

T. Rebernik, B. Halpern, T. Tienkamp, R. Jonkers, A. Noiray, R. van Son, M. van den Brekel, M. Witjes, and M. Wieling
Speech Motor Control 2022
The effect of masking noise on oral cancer speech acoustics and kinematics
ABSTRACT: TBA POSTER

Janay S.C. Monen 1, Bence M. Halpern, Teja Rebernik, Thomas Tienkamp, Rob J.J.H. van Son, Vass Verkhodanova, Max J.H. Witjes,and Martijn Wieling
Young Female Researchers Workshop 2022
Automatic Detection and Severity Estimation for Oral Cancer Speech
ABSTRACT: TBA POSTER

Kirsten Wildenburg, Bence M. Halpern, Teja Rebernik, Thomas Tienkamp, Rob J.J.H. van Son, Vass Verkhodanova, Max J.H. Witjes,and Martijn Wieling
Young Female Researchers Workshop 2022
Automatic Speech Recognition and Error Analyses of Dutch Oral Cancer Speech
ABSTRACT: TBA POSTER

Preprints

Quantifying Bias in Automatic Speech Recognition
Siyuan Feng, Olya Kudina, Bence Mark Halpern, Odette Scharenborg
Paper

Manipulation of oral cancer speech using neural articulatory synthesis
Bence Mark Halpern, Teja Rebernik, Thomas Tienkamp, Rob van Son, Michiel van den Brekel, Martijn Wieling, Max Witjes, Odette Scharenborg
Paper

Doctoral Thesis

Making speech technology accesible for pathological speakers
Supervisors: Michiel van den Brekel, Rob van Son, Odette Scharenborg
PhD Thesis Recording of the defense

Master’s Thesis

Sparse Bayesian Regression for Identification of Gene Regulatory Networks
Supervisors: Dr. Zoltan Tuza, Dr. Wei Pan, Dr. Guy-Bart Stan
Master’s Thesis

Papers where my contribution is acknowledged

Do not waste your electrodes – principles of optimal electrode geometry for spike sorting
Róbert Tóth, Albert Miklós Barth, Andor Domonkos, Viktor Varga, Zoltán Somogyvári
Paper

Computationally efficient neural network classifiers for next generation closed loop neuromodulation therapy – a case study in epilepsy
Ali Kavoosi, Robert Toth, Moaad Benjaber, Mayela Zamora, Antonio Valentın, Andrew Sharott and Timothy Denison
Paper