Dynamic Neural Network-based Solutions for Acoustic Echo Suppression
DOI:
https://doi.org/10.54097/hset.v15i.2795Keywords:
Acoustic Echo Suppression; Non-linear Echo; Dynamic Neural Networks.Abstract
Adaptive filters are widely used as core algorithms in current acoustic echo cancellation (AEC) systems. Echo path estimation is carried out by using different adaptive strategies in the adaptive filter. When nonlinear echo occurs, the performance deteriorates and seriously affects the call quality of both terminals. Taking advantage of the excellent non-linear fitting ability of neural networks in this paper, dynamic neural networks with self-updating parameters work continuously during the inference stage. The algorithm was computed and evaluated using publicly available echo audio data, showing that the dynamic neural network performs approximately as well as the optimal algorithm in the linear echo environment, and outperforms existing algorithms in the non-linear echo environment.
Downloads
References
Gustafsson S, Martin R, Jax P, et al. “A psychoacoustic approach to combined acoustic echo cancellation and noise reduction”. IEEE Transactions on Speech and Audio Processing, 2002, 10(5): 245-256.
Reuven G, Gannot S, Cohen I. “Joint noise reduction and acoustic echo cancellation using the transfer-function generalized sidelobe canceller”. Speech communication, 2007, 49(7-8): 623-635.
Schwarz A, Hofmann C, Kellermann W. “Combined nonlinear echo cancellation and residual echo suppression”. Speech Communication; 11. ITG Symposium. VDE, 2014: 1-4.
Gao Y, Liu I, Zheng J, et al. “Independent Echo Path Modeling for Stereophonic Acoustic Echo Cancellation”. INTERSPEECH. 2020: 3955-3958.
Wu D, Pigou L, Kindermans P J, et al. “Deep dynamic neural networks for multimodal gesture segmentation and recognition”. IEEE transactions on pattern analysis and machine intelligence, 2016, 38(8): 1583-1597.
Han Y, Huang G, Song S, et al. “Dynamic neural networks: A survey”. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021.
Xu Z, Li S, Zhou X, et al. “Dynamic neural networks based kinematic control for redundant manipulators with model uncertainties”. Neurocomputing, 2019, 329: 255-266.
Pfeifenberger L, Pernkopf F. “Nonlinear Residual Echo Suppression Using a Recurrent Neural Network”. Interspeech. 2020: 3950-3954.
Cohen A, Barnov A, Markovich-Golan S, et al. “Joint beamforming and echo cancellation combining QRD based multichannel AEC and MVDR for reducing noise and non-linear echo”. 26th European Signal Processing Conference (EUSIPCO). IEEE, 2018: 6-10.
Ivry A, Cohen I, Berdugo B. “Deep residual echo suppression with a tunable tradeoff between signal distortion and echo suppression”. ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2021: 126-130.
Zhang H, Wang D. “Deep learning for acoustic echo cancellation in noisy and double-talk scenarios”. Training, 2018, 161(2): 322.
Wang Z, Na Y, Liu Z, et al. “Weighted recursive least square filter and neural network based residual echo suppression for the aec-challenge”. ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2021: 141-145.
Zhang S, Wang Z, Sun J, et al. “Multi-task deep residual echo suppression with echo-aware loss”. ICASSP 2022-2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2022: 9127-9131.
Sridhar K, Cutler R, Saabas A, et al. “ICASSP 2021 acoustic echo cancellation challenge: Datasets, testing framework, and results”. ICASSP 2021-2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2021: 151-155.
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
