Briegleb, Annika
Annika Briegleb, M. Sc.
Forschung
Mein Forschungsschwerpunkt liegt auf der Audiosignalverbesserung in diversen Szenarien, u.a. dem robotischen Hören. Dabei untersuche ich hauptsächlich Methoden des maschinellen Lernens und Kombinationen aus modell- und datengetriebenen Verfahren.
Abschlussarbeiten Abgeschlossen/Laufend
Masterarbeiten:
- Complex-valued Variational Autoencoder for Speech Enhancement (2023)
- Dual-staging in speech enhancement: An analysis of cost function modalities (2022)
- Complex-valued Variational Autoencoder for Speech Enhancement (2022)
- Exploring Attention Models for Speech Enhancement (2021)
- Acoustic Source Separation based on Deep Clustering and Independent Component Analysis (2021)
- An Evaluation of the Perception-based Loss for Speech Enhancement (2021)
- A Denoising Autoencoder for Speech Enhancement (2020)
- Deep Attractor Networks for single-channel ego-noise reduction in robot audition (2020)
Bachelorarbeiten:
- Analysis of RNN features in spatiospectral filtering (2024)
- Investigation of the STFT in the context of neural network-based speech enhancement (2022)
- Attention Models for Speech Processing (2020)
Forschungsprojekte:
- Deep learning-based spatial filtering for audio processing (2023)
- Effect of a 3D convolutional layer on multichannel speech enhancement (2023)
- Influence of training target on spatial filtering behavior in multichannel speech enhancement (2023)
- Experimental study on performance variability in neural networks due to hardware and software involved in training (2022)
- Postprocessing for mask-based speech enhancement (2022)
- Evaluation of cost functions for neural network-based postfiltering in acoustic echo cancellation (2021)
- An end-to-end ASR system for speech enhancement (2021)
- Learning a Transformation for Audio Signal Representation (2021)
- Evaluation of Deep Clustering for discriminating various types of robotic ego-noise (2020)
- Hyperparameter adaptation for Deep Clustering for ego-noise suppression (2020)
Publikationen
2023
Exploiting spatial information with the informed complex-valued spatial autoencoder for target speaker extraction
2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) (Rhodes, Greece, 4. Juni 2023 - 10. Juni 2023)
DOI: 10.1109/ICASSP49357.2023.10095196
URL: https://ieeexplore.ieee.org/document/10095196
BibTeX: Download
, , :
Localizing Spatial Information in Neural Spatiospectral Filters
2023 31st European Signal Processing Conference (EUSIPCO) (Helsinki, Finland, 4. September 2023 - 8. September 2023)
DOI: 10.23919/EUSIPCO58844.2023.10289820
URL: https://ieeexplore.ieee.org/document/10289820
BibTeX: Download
, , , :
2022
Statistical Analysis of Randomness in Training of Small-Scale Neural Networks for Speech Enhancement
2022 International Workshop on Acoustic Signal Enhancement (IWAENC) (Bamberg, 5. September 2022 - 8. September 2022)
DOI: 10.1109/IWAENC53105.2022.9914739
URL: https://ieeexplore.ieee.org/document/9914739
BibTeX: Download
, :
2021
Combining Adaptive Filtering and Complex-valued Deep Postfiltering for Acoustic Echo Cancellation
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) (Toronto, 6. Juni 2021 - 11. Juni 2021)
In: IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2021
DOI: 10.1109/ICASSP39728.2021.9414868
BibTeX: Download
, , , , :
2019
Deep Clustering for single-channel ego-noise suppression
International Congress on Acoustics (ICA) (Aachen, 9. September 2019 - 13. September 2019)
URL: https://pub.dega-akustik.de/ICA2019/data/articles/000705.pdf
BibTeX: Download
, , :