DAGA 2025

Kopenhagen / 17. März 2025 - 20. März 2025

DAS I DAGA 2025 – 51. Jahrestagung für Akustik

Das Fraunhofer IDMT nimmt an der DAS I DAGA 2025, der ersten gemeinsamen Jahrestagung der Deutschen Gesellschaft für Akustik (DEGA) und der Dänischen Gesellschaft für Akustik (DAS) als Aussteller und mit mehreren Fachbeiträgen teil.

Am Stand des Ilmenauer Fraunhofer IDMT können sich Besucherinnen und Besucher über aktuelle Forschungsaktivitäten zu den Themen Intelligente akustische Sensorik und Analyse von Industrie- und Umweltgeräuschen informieren. Besuchen Sie uns am Stand B4-34 und tauschen Sie sich mit unseren Wissenschaftlerinnen und Wissenschaftlern aus.

Beiträge des Fraunhofer IDMT zur DAGA 2025

zu den Themen Sound Field Control, Intelligente akustische Sensorik, Analyse von Industrie- und Umweltgeräuschen sowie Sprachverständlichkeit, Sprachwahrnehmung und Sprachsteuerung

Session »Hearing Acoustics«

Real-time assessment of listening effort using non-intrusive binaural prediction models

Referent: Jan Rennies-Hochmuth
Vortrag I 17.03.2025 I 15.40 Uhr I Raum 20

Zum Abstract

Session »Bioacoustics«

Analysing Bowel Sound Signals: Insights from SonicGuard Data Collection

Referent: Danilo Hollosi
Poster I 18.03.2025 I 15.20 Uhr I Nordhavn (B3M10)

Zum Abstract

Session »Bioacoustics«

Detecting Bark Beetle Activity in a Timber Log via Airborne Sound

Referent: Jacopo Barzon
Poster I 18.03.2025 I 15.20 Uhr I Nordhavn (B3M10)

Zum Abstract

Session »Bioacoustics«

Domain Adaptation for Automatic Bird Sound Classification: Compensating for Domain Shift Due to Location Mismatch

Referent: Amir Latifi Bidarouni
Poster I 18.03.2025 I 15.20 Uhr I Nordhavn (B3M10)

Zum Abstract

Session »Psychoacoustics«

Investigation of Training Effects in a Developed Distance Estimation Task Paradigm

Referentin: Carolin Schnetzer

Poster I 18.03.2025 I 15.20 Uhr I Raum 6+7

Zum Abstract

Session »Noise at the Work Place«

Individual Case or a More General Problem? Investigating Altered Localization Cues with Different Level-Dependent Hearing Protection Devices Concerning Worker Safety

Referentin: Sina Buchholz
Vortrag I 19.03.2025 I 11.00 Uhr I Nordhavn (B3M10)

Zum Abstract

Session »Acoustic Scene Analysis using Artificial Intelligence«

Extension of the auditory information in time-frequency representations of audio signals for machine learning applications

Referent: Draško Mašović
Vortrag I 19.03.2025 I 14.20 Uhr I Raum 20

Zum Abstract

Session »Acoustic Scene Analysis using Artificial Intelligence«

Semi-supervised Learning for Acoustic Scene Classification using FixMatch

Referent: Jakob Abeßer
Vortrag I 19.03.2025 I 14.40 Uhr I Raum 20

Zum Abstract

Session »Acoustic Scene Analysis using Artificial Intelligence«

Bridging The Gap: GANs As A Solution For Data-Scarce Industrial Audio Classification

Referentin: Pitchapa Ngamthipwatthana
Poster I 19.03.2025 I 15.20 Uhr I Raum 20

Zum Abstract

Session »Electroacoustics and Audio Signal Processing«

Investigation of Electroacoustic Properties of Experimental Fibre-Optic Microphone Designs

Referent: Florian Klein
Vortrag I 20.03.2025 I 9.00 Uhr I Vesterbro (B3M2)

Zum Abstract

Session »Soundscape«

Variance-maximizing acoustic dimensions for machine learning-based soundscape assessment

Referent: Jakob Bergner
Vortrag I 20.3.2025 I 10.40 Uhr I Sydhavn (B4M4)

Zum Abstract

Session »Soundscape«

Automatic Retrieval of Indicator Sounds for Acoustic Geo-Tagging

Referent: Jakob Abeßer
Poster I 20.03.2025 I 11:00 Uhr I Sydhavn (B4M4)

Zum Abstract

Session »Electroacoustics and Audio Signal Processing«

Modeling of Microchannel Acoustics: Relevant Phenomena and Acoustic Impedance Models

Referent: Andreas Männchen
Poster I 20.03.2025 I 11.00 Uhr I Vesterbro (B3M2)

Zum Abstract

Session »Speech Processing«

Multilingual Listening Effort Prediction - LEAP goes international

Referent: Rainer Huber
Poster I 20.03.2025 I 11.00 Uhr I Raum 19

Zum Abstract

Session »Acoustic Measurements and Sensor Technology«

MPEG Audio Coding for Machines

Referent: Thomas Sporer
Vortrag I 20.3.2025 I 15.40 Uhr I Østerbro (B3M6)

Zum Abstract

Strukturierte Session »Acoustic Scene Analysis using Artificial Intelligence«

Unser Experte Jakob Abeßer organisiert gemeinsam mit Toni Heittola (Universität Tampere, Finnland) die strukturierte Session »Acoustic Scene Analysis using Artificial Intelligence«.

The session explores key research directions in acoustic scene classification, sound event detection and sound event localization and detection to enhance model performance and address existing challenges related to

Robustness to Device or Location Mismatch: Ensuring analysis systems perform well across varying recording devices or locations is crucial for applications like context-aware hearing aids or mobile phones. The focus is on models that generalize effectively in dynamic environments.
Data Scarcity: Handling limited or imbalanced training data is a major challenge. Techniques to improve performance with minimal data are being explored to develop more resilient analysis systems
Relation Between Acoustic Scene Classification and Sound Event Detection: Investigating how specific sound events and their temporal patterns characterize acoustic scenes helps refine acoustic scene classification models, improving classification accuracy and how acoustic scene information can improve sound event detection process.
Efficient Neural Network Structures: Techniques like pruning, model distillation, and student-teacher learning are being applied to create lightweight, efficient models for resource-constrained devices.
Cross-disciplinary Insights: Acoustic scene analysis's relationship to auditory scene analysis, room acoustics, and VR/AR highlights its broader applications, aiming for more accurate and immersive classification systems.

Weitere Beiträge zur DAGA 2025, an denen das Fraunhofer IDMT beteiligt ist

Analysis of Room Acoustic Measurements Indoors, Outdoors and In Between
Referent: Paul Popp (TU Ilmenau)
Poster I 18.03.2025 I 15.20 Uhr I Raum 3

Using interaural mistuning for modelling binaural processing inaccuracies in human speech recognition
Referent: Simon Weihe (CvO Universität Oldenburg)
Vortrag I 18.03.2025 I 16.00 Uhr I Raum 6+7

Evaluating a Binaural Model Using Real-Time Assessment of Listening Effort in Human Listeners
Referent: Martin Berdau (CvO Universität Oldenburg)
Vortrag I 18.03.2025 I 16.20 Uhr I Raum 6+7

Comparison of Noise Detection Methods for Exponential Sine Sweep Measurements
Referent: Georg Stolz (TU Ilmenau)
Vortrag I 19.03.2025 I 9.00 Uhr I Raum 5

Validation of the German Matrix Test in a Remote Testing Application
Referent: Thomas Schwarz (CvO Universität Oldenburg)
Poster I 19.03.2025 I 15.20 Uhr I Raum 6+7

Integrating Multiscale Representation And Re-Evaluating Channel Shuffling In Efficient Time-Frequency Separate Networks For Acoustic Scene Classification
Referent: Yuxuan He (TU Ilmenau)
Poster I 19.03.2025 I 15.20 Uhr I Raum 20

Challenges of conducting a large-scale online perceptual study in the field of spatial audio
Referent: Lukas Treybig (TU Ilmenau)
Vortrag I 20.03.2025 I 14.20 Uhr I Raum 5

NeuroSensEar: Neuromorphic acoustic sensing for the high-performance hearing aids of tomorrow
Referent: Claudia Lenk (Universität Ulm)
Vortrag I 20.03.2025 I 14.40 Uhr I Vesterbro (B3M2)

Comparison of reference-based and single-ended prediction models for perceived listening effort
Referent: Jan Reimes (HEAD acoustics GmbH)
Poster I 20.03.2025 I 15.00 Uhr I Raum 19

Insights Into Automated Quality Inspection in Lightweight Additive Manufacturing Using Machine-Learning With Measured and Simulated Acoustic Emissions
Referent: Sören Keuchel (Novicos GmbH)
Vortrag I 20.03.2025 I 15.20 Uhr I Raum 18

Fachtagung

Kopenhagen / 17. März 2025 - 20. März 2025

DAS I DAGA 2025 – 51. Jahrestagung für Akustik

Beiträge des Fraunhofer IDMT zur DAGA 2025

Real-time assessment of listening effort using non-intrusive binaural prediction models

Analysing Bowel Sound Signals: Insights from SonicGuard Data Collection

Detecting Bark Beetle Activity in a Timber Log via Airborne Sound

Domain Adaptation for Automatic Bird Sound Classification: Compensating for Domain Shift Due to Location Mismatch

Investigation of Training Effects in a Developed Distance Estimation Task Paradigm

Individual Case or a More General Problem? Investigating Altered Localization Cues with Different Level-Dependent Hearing Protection Devices Concerning Worker Safety

Extension of the auditory information in time-frequency representations of audio signals for machine learning applications

Semi-supervised Learning for Acoustic Scene Classification using FixMatch

Bridging The Gap: GANs As A Solution For Data-Scarce Industrial Audio Classification

Investigation of Electroacoustic Properties of Experimental Fibre-Optic Microphone Designs

Variance-maximizing acoustic dimensions for machine learning-based soundscape assessment

Automatic Retrieval of Indicator Sounds for Acoustic Geo-Tagging

Modeling of Microchannel Acoustics: Relevant Phenomena and Acoustic Impedance Models

Multilingual Listening Effort Prediction - LEAP goes international

MPEG Audio Coding for Machines

Strukturierte Session »Acoustic Scene Analysis using Artificial Intelligence«

Weitere Beiträge zur DAGA 2025, an denen das Fraunhofer IDMT beteiligt ist

Contact Press / Media

Christoph Sladeczek

Contact Press / Media

PD Dr. Jan Rennies-Hochmuth