ICASSP 2021 Acoustic Echo Cancellation Challenge: Datasets, Testing Framework, and Results

Sridhar, Kusha; Cutler, Ross; Saabas, Ando; Pärnamaa, Tanel; Loide, Markus; Gamper, Hannes; Braun, Sebastian; Aichner, Robert; Subramaniam, Sriram

doi:10.48550/arxiv.2009.04972

Cited by 8 publications

(22 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The RES thus requires 4.6% of an x86 mobile CPU core (Intel i7-8565U) to operate in real-time. When combined with the AEC, the total complexity of the proposed 16 kHz echo control solution as submitted to the AEC challenge [17] is 5.5% CPU (0.55 ms per 10-ms frame). Since the RES is already designed to operate at 48 kHz, the total cost of fullband echo control only increases to 6.6%, with the difference due to the increased AEC sampling rate.…”

Section: Experiments and Resultsmentioning

confidence: 99%

“…Our model is trained (Section 5) to enhance the speech from the AEC using the far-end signal as side information to help remove the far-end signal while denoising the near-end speech. Results from our experiments and from the Acoustic Echo Cancellation Challenge [17] show that the proposed algorithm outperforms both traditional and other neural approaches to residual echo suppression, taking first place in the challenge (Section 6).…”

Section: Introductionmentioning

confidence: 86%

“…The convolutional layers are aligned in time such as to use up to M frames into the future. In order to achieve the 40 ms algorithmic delay allowed by the challenge [17], including the 10-ms frame size and the 10-ms overlap, we have M = 2.…”

Section: Dnn Modelmentioning

confidence: 99%

See 2 more Smart Citations

Low-Complexity, Real-Time Joint Neural Echo Control and Speech Enhancement Based On PercepNet

Valin¹,

Tenneti²,

Helwani³

et al. 2021

Preprint

View full text Add to dashboard Cite

Speech enhancement algorithms based on deep learning have greatly surpassed their traditional counterparts and are now being considered for the task of removing acoustic echo from hands-free communication systems. This is a challenging problem due to both real-world constraints like loudspeaker non-linearities, and to limited compute capabilities in some communication systems. In this work, we propose a system combining a traditional acoustic echo canceller, and a low-complexity joint residual echo and noise suppressor based on a hybrid signal processing/deep neural network (DSP/DNN) approach. We show that the proposed system outperforms both traditional and other neural approaches, while requiring only 5.5% CPU for real-time operation. We further show that the system can scale to even lower complexity levels.

show abstract

Section: Experiments and Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 86%

See 1 more Smart Citation

Low-Complexity, Real-Time Joint Neural Echo Control and Speech Enhancement Based On PercepNet

Valin¹,

Tenneti²,

Helwani³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…We train and test our optimizers using the synthetic portion of the Microsoft AEC Challenge dataset [30]. This dataset includes farend noise, near-end noise, and far-end nonlinearities.…”

Section: Datasetmentioning

confidence: 99%

“…To demonstrate our approach, we learn to optimize an AEC task as shown in Fig.1for a single-talk in noise scenario. We use the Microsoft AEC Challenge dataset[30] to learn update rules for a variety of common linear and nonlinear multidelayed block frequency domain filters (MDF)[31]. We compare our results to handengineered, grid-search-tuned block NLMS and RMSprop[32] optimizers, as well as the open-source Speex AEC[8,33].…”

mentioning

confidence: 99%

Auto-DSP: Learning to Optimize Acoustic Echo Cancellers

Casebeer¹,

Bryan²,

Smaragdis³

2021

Preprint

View full text Add to dashboard Cite

Adaptive filtering algorithms are commonplace in signal processing and have wide-ranging applications from single-channel denoising to multi-channel acoustic echo cancellation and adaptive beamforming. Such algorithms typically operate via specialized online, iterative optimization methods and have achieved tremendous success, but require expert knowledge, are slow to develop, and are difficult to customize. In our work, we present a new method to automatically learn adaptive filtering update rules directly from data. To do so, we frame adaptive filtering as a differentiable operator and train a learned optimizer to output a gradient descent-based update rule from data via backpropagation through time. We demonstrate our general approach on an acoustic echo cancellation task (single-talk with noise) and show that we can learn high-performing adaptive filters for a variety of common linear and non-linear multidelayed block frequency domain filter architectures. We also find that our learned update rules exhibit fast convergence, can optimize in the presence of nonlinearities, and are robust to acoustic scene changes despite never encountering any during training.

show abstract

Acoustic Echo Cancellation with the Dual-Signal Transformation LSTM Network

Westhausen

Meyer

2021

ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

This paper applies the dual-signal transformation LSTM network (DTLN) to the task of real-time acoustic echo cancellation (AEC). The DTLN combines a short-time Fourier transform and a learned feature representation in a stacked network approach, which enables robust information processing in the time-frequency and in the time domain, which also includes phase information. The model is only trained on 60 h of real and synthetic echo scenarios. The training setup includes multi-lingual speech, data augmentation, additional noise and reverberation to create a model that should generalize well to a large variety of real-world conditions. The DTLN approach produces state-of-the-art performance on clean and noisy echo conditions reducing acoustic echo and additional noise robustly. The method outperforms the AEC-Challenge baseline by 0.30 in terms of Mean Opinion Score (MOS).

show abstract

ICASSP 2021 Acoustic Echo Cancellation Challenge: Datasets, Testing Framework, and Results

Cited by 8 publications

References 0 publications

Low-Complexity, Real-Time Joint Neural Echo Control and Speech Enhancement Based On PercepNet

Low-Complexity, Real-Time Joint Neural Echo Control and Speech Enhancement Based On PercepNet

Auto-DSP: Learning to Optimize Acoustic Echo Cancellers

Acoustic Echo Cancellation with the Dual-Signal Transformation LSTM Network

Contact Info

Product

Resources

About