Navigation

Programm

Die Programmübersicht finden Sie hier.

Mittwoch - 10.10.2018

Gebäude W32
14:00-14:10

Eröffnung und Begrüßung
Leitung: Simon Doclo, Peter Jax

Grußwort der Vizepräsidentin für wissenschaftlichen Nachwuchs und Internationales
Prof. Dr. Esther Ruigendijk

Gebäude W32
14:10-15:10
Prof. Dr. Sharon Gannot, Bar-Ilan University, Israel (Keynote)
Moderation: Simon Doclo

Multi-Microphone Speaker Localization and Tracking on Manifolds
 
Gebäude W32
15:10-16:30
Active Noise Cancellation (Tutorial Session)
Leitung: Peter Jax, Peter Vary

Signal Processing Challenges for Active Noise Cancellation Headphones
Stefan Liebich, Johannes Fabry, Peter Jax and Peter Vary

Hybrid Active Noise Control Structures: A Short Overview
Piero Rivera Benois, Patrick Nowak and Udo Zölzer

Active control of noise and vibration for time-varying transfer paths
Arthur Berkhoff

The Occlusion Effect - Measurements, Simulations, and Countermeasures
Thomas Zurbrügg
Gebäude W30
16:30-17:00
Kaffeepause
Gebäude W30
17:00-18:30

Poster- und Show&Tell-Sitzung
Leitung: Emanuël Habets

P1: Online Adaptation for Jointly Optimized Blind Source Separation and Dereverberation of Speech Mixtures
Timo Schuster and Stefan Feldes

P2: Subband Based Room Impulse Response Reshaping
Radoslaw Mazur, Fabrice Katzberg, Martina Böhme and Alfred Mertins

P3: Good Noise Power Estimators Are Not Always Good
Patrick Meyer and Samy Elshamy

P4: Evaluation of Robust Constrained MFMVDR Filtering for Single-Channel Speech Enhancement
Dörte Fischer and Simon Doclo

P5: Evaluation of binaural Own Voice Detection (OVD) algorithms
Joerg Bitzer, Sascha Bilert and Inga Holube

P6: Distributed MAP Estimators for Noise Reduction in Fully Connected Wireless Acoustic Sensor Networks
Raziyeh Ranjbaryan, Simon Doclo and Hamid Reza Abutalebi

P7: Benchmarking Neural Network Architectures for Acoustic Sensor Networks
Janek Ebbers, Jens Heitkaemper, Joerg Schmalenstroeer and Reinhold Haeb-Umbach

P8: Unsupervised Domain Adaptation by Adversarial Learning for Robust Speech Recognition
Pavel Denisov, Ngoc Thang Vu and Marc Ferras Font

P9: Parameter Optimization for CTC Acoustic Models in a Less-resourced Scenario: An Empirical Study
Markus Müller, Sebastian Stüker and Alex Waibel

P10: Keyword Detection for the Activation of Speech Assistants
Hans-Günter Hirsch and Michael Gref

P11: Utilizing Slow Feature Analysis for Lipreading
Jan Freiwald, Mahdie Karbasi, Steffen Zeiler, Jan Melchior, Varun Kompella, Laurenz Wiskott and Dorothea Kolossa

P12: Diagnostic and Summative Approach for Predicting Speech Communication Quality in a Super-Wideband Context
Sebastian Möller, Tobias Hübschen, Gabriel Mittag and Gerhard Schmidt

P13: Observations from a Simple Vocal-Tract-Model's Behaviour for PD-Dysarthric Speech: Applicability
Ulrich Heute and Gerhard Schmidt

P14: The Articulatory Code and Related OVC-Gestures
Harald Höge

Demonstratoren

3PASS - Background noise testing matters
Christian Schüring

Binaural steering beamformer hearing aid
Kamil Adiloglu, Reza Verzandeh and Hendrik Kayser

Open-source, mobile, realtime hearing aid research platform
Paul Maanen, Tobias Herzke, Hendrik Kayser and Volker Hohmann

Natural own-voice perception while wearing headphones - Active Occlusion Cancellation with Equalized Hear-Through
Stefan Liebich, Lars Thieling, Johannes Fabry, Peter Jax and Peter Vary

Gebäude W33
18:00-19:00
Sitzung der ITG-Fachausschüsse AT3 und AT4
19:30

Welcome Reception
Bar und Restaurant Patio
Ort: Bahnhofstraße 11, Oldenburg

 

 

Donnerstag - 11.10.2018

Gebäude W32
08:30-09:30
Prof. Dr. Shinji Watanabe, Johns Hopkins University, USA (Keynote)
Moderation: Simon Doclo

Neural End-to-End Architectures for Speech Recognition in Adverse Environments
 
Gebäude W32
09:30-10:30
Signal Enhancement for Hearing Aids and Communication Systems (Lecture Session)
Leitung: Heinrich Löllmann, Rainer Martin

A Relative Transfer Function-based Post-Filter for Speech Enhancement in Hearing Aids using a Nearby External Microphone
Dianna Yee, Homayoun Kamkar-Parsi, Rainer Martin and Henning Puder

Multi-loudspeaker equalization for acoustic transparency in a custom hearing device
Henning Schepker, Florian Denk, Birger Kollmeier and Simon Doclo

On the Benefit of a Stereo Acoustic Echo Cancellation in an In-Car Communication System
Jan Franzen, Inka Meyer Zum Alten Borgloh and Tim Fingscheidt

Acoustic Tube Interpolation for Spectral Envelope Estimation in Artificial Bandwidth Extension
Thomas Schlien, Peter Jax and Peter Vary
Gebäude W30
10:30-11:00
Kaffeepause
Gebäude W30
11:00-12:30

Poster- und Show&Tell-Sitzung
Leitung: Dorothea Kolossa

P1: Enhancement of G.711-Coded Speech Providing Quality Higher Than Uncoded
Ziyue Zhao, Huijun Liu and Tim Fingscheidt

P2: NARA-WPE: A Python package for weighted prediction error dereverberation in Numpy and Tensorflow for online and offline processing
Lukas Drude, Jahn Heymann, Christoph Boeddeker and Reinhold Haeb-Umbach

P3: Iterative Alternating Least-Squares Approach to Jointly Estimate the RETFs and the Diffuse PSD
Marvin Tammen, Ina Kodrasi and Simon Doclo

P4: Equalization filter design for achieving acoustic transparency in a semi-open fit hearing device
Florian Denk, Henning Schepker, Simon Doclo and Birger Kollmeier

P5: A gaze-based attention model for spatially-aware hearing aids
Giso Grimm, Hendrik Kayser, Maartje Hendrikse and Volker Hohmann

P6: Evaluation of Signal-Dependent Partial Noise Estimation Algorithms for Binaural Hearing Aids
Jonas Klug, Daniel Marquardt, Nico Gößling and Simon Doclo

P7: Maximum-Likelihood and Maximum-A-Posteriori Perspectives for Blind Channel Identification on Acoustic Sensor Network Data
Philipp Thüne and Gerald Enzner

P8: Insights into the Interplay of Sampling Rate Offsets and MVDR Beamforming
Joerg Schmalenstroeer and Reinhold Haeb-Umbach

P9: Open Source Automatic Speech Recognition for German
Benjamin Milde and Arne Köhn

P10: Improving Robust Speech Recognition for German Oral History Interviews Using Multi-Condition Training
Michael Gref, Christoph Schmidt and Joachim Köhler

P11: Robust Speaker Identification by Fusing Classification Scores with a Neural Network
Kevin Wilkinghoff, Paul M. Baggenstoss, Alessia Cornaggia-Urrigshardt and Frank Kurth

P12: Super-Wideband Extension of a Perceptual Based Echo Assessment Method for Aurally Adequate Evaluation of Residual Single Talk Echoes
Stefan Bleiholder, Jan Reimes and Frank Kettler

P13: Statistical Modeling of Speech Spectral Coefficients in Patients with Parkinson's Disease
Ina Kodrasi and Hervé Bourlard

P14: Session-Independent Array-Based EMG-to-Speech Conversion using Convolutional Neural Networks
Lorenz Diener, Gerrit Felsch, Miguel Angrick and Tanja Schultz

Demonstratoren

3PASS - Background noise testing matters
Christian Schüring

Binaural steering beamformer hearing aid
Kamil Adiloglu, Reza Verzandeh and Hendrik Kayser

A gaze-based attention model for hearing device control -- real-time demonstrator
Giso Grimm, Hendrik Kayser, Frederike Kirschner and Volker Hohmann

The acoustically transparent hearing device
Henning Schepker, Florian Denk, Birger Kollmeier and Simon Doclo

Human Dialogue Simulation Framework for the Estimation of Conversational Quality
Thilo Michael

Virtual Reality Lab
Jan Heeren and Stephan Ewert 

Gebäude W30
12:30-13:30

Mittagspause

Gebäude W32
13:30-14:30
Diagnostics, Paralinguistics, and Biosignals (Lecture Session)
Leitung: Tanja Schultz, Björn Schuller

Automatic Screening for Transition into Dementia using Speech
Jochen Weiner and Tanja Schultz

Evaluation of the Pain Level from Speech: Introducing a Novel Pain Database and Benchmarks
Zhao Ren, Nicholas Cummins, Jing Han, Sebastian Schnieder, Jarek Krajewski and Björn Schuller

On the Effects of Speaker Gender in Emotion Recognition Training Data
Ziyi Xu, Patrick Meyer and Tim Fingscheidt

A comparison of EMG-to-Speech Conversion for Isolated and Continuous Speech
Lorenz Diener, Sebastian Bredehöft and Tanja Schultz
Gebäude W32
14:30-15:30
Neural Processing in Automatic Speech Recognition (Lecture Session)
Leitung: Bernd Meyer, Ralf Schlüter

DNN/CNN Acoustic Model Turbo Fusion for Phoneme Recognition
Timo Lohrenz, Wei Li and Tim Fingscheidt

Sequence Modeling and Alignment for LVCSR-Systems
Eugen Beck, Albert Zeyer, Patrick Doetsch, André Merboldt, Ralf Schlüter and Hermann Ney

Accurately Capturing Speech Feature Distributions by Extending Supervectors for Robust Speaker Recognition
Kevin Wilkinghoff

Objective Assessment of a Speech Enhancement Scheme with an Automatic Speech Recognition-Based System
Rainer Huber, Arne Pusch, Niko Moritz, Jan Rennies, Henning Schepker and Bernd Meyer
Gebäude W32
15:30-16:00
Kaffeepause
16:30-18:00 Stadtführung in Oldenburg
Ort: Schlossplatz 27, Oldenburg
19:30 Abendveranstaltung im Oldenburger Schloss
Ort: Schlossplatz 27, Oldenburg

 

Freitag - 12.10.2018

Gebäude W32
09:00-10:00
Dr. Thomas Brand, University of Oldenburg, Germany (Keynote)
Moderation: Peter Jax

Modeling human speech recognition
 
Gebäude W32
10:00-11:00
Acoustic Sensor Networks (Lecture Session)
Leitung: Gerald Enzner, Jörg Schmalenströer

Resource Allocation for Distributed Blind Source Separation
Markus Bachmann, Andreas Brendel and Walter Kellermann

Fast and Accurate Audio Resampling for Acoustic Sensor Networks by Polyphase-Farrow Filters with FFT Realization
Aleksej Chinaev, Gerald Enzner and Joerg Schmalenstroeer

Gender Discrimination Versus Speaker Identification Through Privacy-Aware Adversarial Feature Extraction
Alexandru Nelus and Rainer Martin

RTF-Based Binaural MVDR Beamformer Exploiting an External Microphone in a Diffuse Noise Field
Nico Gößling and Simon Doclo
Gebäude W30
11:00-11:30
Kaffeepause
Gebäude W30
11:30-13:00

Poster- und Show&Tell-Sitzung
Leitung: Tim Fingscheidt

P1: Optimal temporal dynamics of MFCCs for low-complexity VAD Systems — a case study
Alexandra Craciun and Tom Bäckström

P2: Overcoming Covariance Matrix Phase Sensitivity in Single-Channel Speech Enhancement with Correlated Spectral Components
Johannes Stahl, Sean Wood and Pejman Mowlaee

P3: Acoustic Howling Detection and Suppression for IP-Based Teleconference Systems
Stefan Kühl, Carlotta Anemüller, Christiane Antweiler, Peter Jax, Florian Heese and Patrick Vicinus

P4: A Synchronized Binaural N-of-M Sound Coding Strategy for Bilateral Cochlear Implant Users
Tom Gajecki and Waldo Nogueira

P5: Performance Prediction of the Binaural MVDR Beamformer with Partial Noise Estimation using a Binaural Speech Intelligibility Model
Christopher Hauth, Nico Gößling and Thomas Brand

P6: Extension of the Framework for auditory discrimination experiments (FADE) to predict the Göttingen (everyday) sentence speech test
David Hülsmeier, Anna Warzybok and Marc René Schädler

P7: MARVELO – A Framework for Signal Processing in Wireless Acoustic Sensor Networks
Haitham Afifi, Joerg Schmalenstroeer, Joerg Ullmann, Reinhold Haeb-Umbach and Holger Karl

P8: Source separation by fuzzy-membership value aware beamforming and masking in ad hoc arrays
Sebastian Gergen, Rainer Martin and Nilesh Madhu

P9: Densely Connected Convolutional Networks for Speech Recognition
Chia-Yu Li and Ngoc Thang Vu

P10: "Gap after the next two vehicles": A Spatio-temporally Situated Dialog for a Cooperative Driving Assistant
Martin Heckmann, Dennis Orth and Dorothea Kolossa

P11: Deep Learning of Articulatory-Based Representations and Applications for Improving Dysarthric Speech Recognition
Feifei Xiong, Jon Barker and Heidi Christensen

P12: Automatic Estimation of the Triangular Vowel Space Area from i-Vectors
Maureen Tanuadji, Michael Stadtschnitzer, Rolf Bardeli and Hagen Jaeger

P13: ANN-based Alzheimer’s disease classification from bag of words
Philipp Klumpp, Julian Fritsch and Elmar Nöth

P14: Optical force and distance sensing in intraoral devices for stroke rehabilitation: a distance calibration and force classification approach
Christoph Wagner, Simon Stone and Peter Birkholz

Demonstratoren

3PASS -- Background noise testing matters
Christian Schüring

Binaural steering beamformer hearing aid
Kamil Adiloglu, Reza Varzandeh and Hendrik Kayser

A gaze-based attention model for hearing device control -- real-time demonstrator
Giso Grimm, Hendrik Kayser, Frederike Kirschner and Volker Hohmann

Accessible infrastructure for hearing research: An almost commodity hardware based mobile prototype of a hearing aid featuring the openMHA.org algorithm research platform
Marc-René Schädler

Speech based home automation
Hans-Günter Hirsch

Virtual Reality Lab
Jan Heeren and Stephan Ewert

RTF-Based Binaural MVDR Beamformer Exploiting an External Microphone
Nico Gößling and Simon Doclo

Gebäude W30
13:00-14:00

Mittagspause

Gebäude W32
14:00-14:45

Speech and Audio Quality Evaluation (Lecture Session)
Leitung: Hans-Wilhelm Gierlich, Sebastian Möller, Gerhard Schmidt

Auditory Assessment of Multichannel Audio Systems
Magnus Schäfer

Signal-based Root Cause Analysis of Quality Impairments in Speech Communication Networks
Tobias Hübschen, Gabriel Mittag, Sebastian Möller and Gerhard Schmidt

Perceived Listening Effort for In-car Communication Systems
Jan Reimes and Christian Lüke

Gebäude W32
14:45-15:30
Machine Learning for Speech Enhancement (Lecture Session)
Leitung: Timo Gerkmann, Reinhold Häb-Umbach

Robust DNN-Based Speech Enhancement with Limited Training Data
Robert Rehr and Timo Gerkmann

Smoothing along Frequency in Online Neural Network Supported Acoustic Beamforming
Jens Heitkaemper, Jahn Heymann and Reinhold Haeb-Umbach

Multichannel Nonnegative Matrix Factorization for Ego-Noise Suppression
Thomas Haubner, Alexander Schmidt and Walter Kellermann
Gebäude W32
15:30-15:45
Tagungsabschluss
Leitung: Simon Doclo, Peter Jax

 

 

 

 

ITG-Wkyebmaster: Sy0l3dimon Docqmnvplox+w (simon.docuyvmolo@uol.degcr) (Stand: 07.11.2019)