The INTERSPEECH 2020 Deep Noise Suppression Challenge: Datasets, Subjective Speech Quality and Testing Framework

By Chandan K. A. Reddy, Ebrahim Beyrami, Harishchandra Dubey, Vishak Gopal, R. Cheng, Ross Cutler, Sergiy Matusevych, Robert Aichner, Ashkan Aazami, Sebastian Braun and 3 more. 2020

91 citations recorded by the source corpus.

The INTERSPEECH 2020 Deep Noise Suppression Challenge is intended to promote collaborative research in real-time single-channel Speech Enhancement aimed to maximize the subjective (perceptual) quality of the enhanced speech. A typical approach to evaluate the noise suppression methods is to use objective metrics on the test set obtained by splitting the original dataset. Many publications report reasonable performance on the synthetic test set drawn from the same distribution as that of the training set. However, often the model performance degrades significantly on real recordings. Also, most of the conventional objective metrics do not correlate well with subjective tests and lab subjective tests are not scalable for a large test set. In this challenge, we open-source a large clean speech and noise corpus for training the noise suppression models and a representative test set to real-world scenarios consisting of both synthetic and real recordings. We also open source an online subjective test framework based on ITU-T P.808 for researchers to quickly test their developments. The winners of this challenge will be selected based on subjective evaluation on a representative test set using P.808 framework.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Real Time Speech Enhancement in the Waveform Domain (2020, methodology): “Training We train the DEMUCS model for 400 epochs on the Valentini [18] dataset and 250 on the DNS [19] dataset.”
Phase-aware Single-stage Speech Denoising and Dereverberation with U-Net (2020, methodology): “We used the DNS challenge dataset [22] and internally collected dataset for training.”
Data Augmentation and Loss Normalization for Deep Noise Suppression (2020, background): “Speech enhancement using neural networks has recently seen large attention and success in research [1,2] and is being implemented in commercial applications also targeting real-time communication.”
Exploring the Best Loss Function for DNN-Based Low-latency Speech Enhancement with Temporal Convolutional Networks (2020, methodology): “Third, we extend our best scheme that was found on the VBD dataset for the DNS Challenge and our modified Conv-TasNet, which we submitted to the DSN Challenge, achieves the best performance.”
Harmonic Gated Compensation Network Plus for ICASSP 2022 DNS Challenge (2022, methodology): “Finally, we add a high-band module to help the model handle the FB signal, resulting in HGCN+. HGCN+ outperforms the referenced models on the DNS-2020 and DNS-2022 test sets.”
Sequence to Multi-Sequence Learning via Conditional Chain Mapping for Mixture Signals (2020, methodology): “To evaluate this, we conduct the speech denoising task based on a recently published dataset from the DNS-Challenge 2020 [29], which consists of 60,000 no-reverberant noisy clips in training and 300 in evaluation set.”
Fullsubnet: A Full-Band and Sub-Band Fusion Model for Real-Time Single-Channel Speech Enhancement (2020, methodology): “We evaluated the FullSubNet on the DNS Challenge (INTER-SPEECH 2020) dataset [16].”