Author: Mads Græsbøll Christensen

Liming Shi defends his Ph.D.

On January 21 2020, Audio Analysis Lab member Liming Shi sucecssfully defended his PhD thesis entitled Speech Modeling and Robust Estimation for Diagnosis of Parkinson’s Disease. The thesis is concerned with new models and methods for diagnosing Parkinson’s disease from voice signals recorded under noisy conditions, and Liming worked the project for three years. The assessment committee was comprised of Prof. Jonathon Chambers (University of Leicester), Dr. Phil Garner (IDIAP), and Assoc. Prof. Troels Pedersen (AAU). Liming Shi was supervised by Prof. Mads Græsbøll Christensen and co-supervised by Assoc. Prof. Jesper Rindom Jensen and Assoc. Prof. Jesper Kjær Nielsen.

ICASSP 2019

Last week, the Audio Analysis Lab presented five papers and a demo at the 44th International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2019, in Brighton, United Kingdom. These also include two contributions from the project:

  • A STUDY ON HOW PRE-WHITENING INFLUENCES FUNDAMENTAL FREQUENCY ESTIMATION by Alfredo Esquivel Jaramillo, Jesper Kjær Nielsen, Mads Græsbøll Christensen
  • QUALITY CONTROL OF VOICE RECORDINGS IN REMOTE PARKINSON’S DISEASE MONITORING USING THE INFINITE HIDDEN MARKOV MODEL by Amir Hossein Poorjam, Yordan P. Raykov, Reham Badawy, Jesper Rindom Jensen, Mads Græsbøll Christensen, Max A. Little

The lab also presented a demo on REAL-TIME BAYESIAN PITCH TRACKING in by Liming Shi, Jesper Kjaer Nielsen, Mads, Graesboll Christensen in the Show & Tell Demonstrations which was developed as part of the project.

Talk at Clinical Science and Engineering for Digital Health Workshop 2018

On August 29, Professor Mads Græsbøll Christensen of the Audio Analysis Lab gave an invited talk entitled “Diagnosis of Parkinson’s Disease from the Voice: Pre-Processing, Robust Estimation, or Noisy Features?” and participated in a panel discussion at the Clinical Science and Engineering for Digital Health Workshop 2018 organized by Aston University in Birmingham, UK. The talk was concerned with how to best deal with signal degradations such as noise in biomedical engineering applications.

Papers Presented at ICASSP 2018

The 43rd IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2018 will be held April 15-20, 2018 in Calgary, Canada. As usual, the Audio Analysis Lab will be present at the top signal processing conference in the world with 8 papers in total, including several from the project. The following papers from the project will be presented at the conference, which is the top signal processing conference in the world:

  • A PARAMETRIC APPROACH FOR CLASSIFICATION OF DISTORTIONS IN PATHOLOGICAL VOICES
  • MULTIPITCH ESTIMATION USING BLOCK SPARSE BAYESIAN LEARNING AND INTRA-BLOCK CLUSTERING
  • A SUPERVISED APPROACH TO GLOBAL SIGNAL-TO-NOISE RATIO ESTIMATION FOR WHISPERED AND PATHOLOGICAL VOICES

The papers were written by Ph.D. students Amir Poorjam and Liming Shi in collaboration with their supervisors.

Audio Analysis Lab members Assistant Prof. Jesper Rindom Jensen, Assistant Prof. Jesper Kjær Nielsen, and Prof. Mads Græsbøll Christensen also presented a tutorial entitled Model-based Speech and Audio Processing at the conference. It includes several highlights from the project.

EliteForsk Travel Stipend for Liming Shi

Audio Analysis Lab member and project employee Liming shi received a prestigious EliteForsk (Elite Research) Travel Stipend from the Ministry of Education and Science yesterday during the annual EliteForsk celebration at the beautiful Copenhagen Opera House. The stipends, which are awarded to current Ph.D. students in Denmark, consists of 200,000 DKK for visiting universities abroad and conference participation. The travel stipends are awarded by nomination and the candiates are evaluated by council members of Independent Research Fund Denmark. Only 20 of these highly competitive stipends are awarded annually, and only two were awarded to Ph.D. students at AAU. Liming Shi will be visiting Cambridge University, United Kingdom and Alto University, Finland. You can read more about the EliteForsk celebration here and about Liming and his work here.

Paper Presented at WASPAA 2017

The Audio Analysis Lab, represented by Liming Shi, presented a paper on Parkinsons at the IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, which was held October 15-18 at Mohonk Mountain House, New Paltz, New York. The lab was also part of the organizing team for this edition of the workshop, as Mads Græsbøll Christensen was Technical Program Co-Chair. The paper was:

A Kalman-Based Fundamental Frequency Estimation Algorithm Liming Shi, Jesper Kjær Nielsen and Jesper Rindom Jensen (Aalborg University, Denmark); Max Little (MIT, USA); Mads Græsbøll Christensen (Aalborg University, Denmark)

Fundamental frequency estimation is an important task in speech and audio analysis. Harmonic model-based methods typically have superior estimation accuracy. However, such methods usually assume that the fundamental frequency and amplitudes are stationary over a short time frame. In this paper, we propose a Kalman filter-based fundamental frequency estimation algorithm using the harmonic model, where the fundamental frequency and amplitudes can be truly nonstationary by modeling their time variations as first-order Markov chains. The Kalman observation equation is derived from the harmonic model and formulated as a compact nonlinear matrix form, which is further used to derive an extended Kalman filter. Detailed and continuous fundamental frequency and amplitude estimates for speech, the sustained vowel /a/ and solo musical tones with vibrato are demonstrated.

Paper Presented at EUSIPCO 2017

A paper from the project has been presented at EUSIPCO 2017 on Kos, Greece. The paper is entitled A Variational EM Method for Pole-Zero Modeling of Speech with Mixed Block Sparse and Gaussian Excitation and was written by Liming Shi, Jesper Kjær Nielsen, Jesper Rindom Jensen and Mads Græsbøll Christensen. Liming Shi presented the paper at the conference. EUSIPCO is the flagship conference of the European Association for Signal Processing (EURASIP).

 

Paper Presented at Interspeech 2017

Project employee and Ph.D. student Amir Poorjam presented a paper entitled Dominant Distortion Classification for Pre-Processing of Vowels in Remote Biomedical Voice Analysis at Interspeech 2017 in Stockholm on August 21 2017. The paper was presented as part of a session on Pathological Speech and Language and was co-authored by Max Little, Jesper Rindom Jensen, and Mads Græsbøll Christensen, all of whom are involved in the Parkinsons project.

Abstract: Advances in speech signal analysis facilitates the development of techniques for remote biomedical voice assessment. However, the performance of these techniques are affected by noise
and distortion in signals. In this paper, we focus on the vowel /a/ as the widely-used voice signal for pathological voice assessments and investigate the impact of four major types of distortion that are commonly present during recording or transmission in voice analysis, namely background noise, reverberation, peak clipping and compression on Mel-frequency cepstral coefficients (MFCCs) as the most popular features in many voice assessments. Then, we propose a new distortion classification
approach to detect the most dominant distortion in such voice signals. The proposed method involves MFCCs as frame-level features and a support vector machine as the classifier to detect the presence and type of distortion in frames of a given voice signal. Experimental results over the healthy and Parkinson’s voices show the effectiveness of the proposed approach in distortion detection.