Acoustic Echo Cancellation with the Dual-Signal Transformation LSTM Network
TL
TL;DR
The DTLN approach produces state-of-the-art performance on clean and noisy echo conditions reducing acoustic echo and additional noise robustly and outperforms the AEC-Challenge baseline by 0.30 in terms of Mean Opinion Score (MOS).
Abstract
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).
References36 items
1
ICASSP 2021 Deep Noise Suppression Challenge
2
ICASSP 2021 Acoustic Echo Cancellation Challenge: Datasets and Testing Framework
3
Data Augmentation and Loss Normalization for Deep Noise Suppression
4
PoCoNet: Better Speech Enhancement with Frequency-Positional Embeddings, Semi-Supervised Conversational Data, and Biased Loss
5