The Audio Degradation Toolbox and Its Application to Robustness Evaluation

By Matthias Mauch, Sebastian Ewert. 2013

85 citations recorded by the source corpus.

Subfields: Safety, Robustness & Interpretability, Speech & Audio

How researchers see this paper

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

In Safety, Robustness & Interpretability

In Speech & Audio

Citations in the corpus

Robust Speech Recognition via Large-Scale Weak Supervision (2022, methodology): “We tested the noise robustness of Whisper models and 14 LibriSpeech-trained models by measuring the WER when either white noise or pub noise from the Audio Degradation Toolbox (Mauch & Ewert, 2013) was added to the audio.”
Sound Event Detection in Synthetic Domestic Environments (2020, unknown): “Six alternative versions of the subset fbsnr 30dB are generated introducing artiﬁcial degradation with the Audio Degradation Toolbox [13].”
Open-Source Practices for Music Signal Processing Research: Recommendations for Transparent, Sustainable, and Reproducible Audio Research (2019, methodology): “Relatedly, degradation methods apply similar techniques to evaluation data as a means of diagnosing failure modes in a model once its parameters have been estimated [17].”
Sound Event Detection in Domestic Environments with Weakly Labeled Data and Soundscape Synthesis (2019, unknown): “…are generated varying the value of the foreground events’ SNR with respect to the background: 0 dB, 6 dB, 15 dB and 30 dB. Audio degradation : Six alternative versions of the previous subset (with SNR=0 dB) are generated introducing artiﬁcial degradation with the Audio Degradation Toolbox [20].”
PYIN: A fundamental frequency estimator using probabilistic threshold distributions (2014, methodology): “In order to simulate a more realistic situation without such clean data, we degraded the audio using five presets of the Audio Degradation Toolbox (ADT [13]) in addition to the original wav files.”
A Unified Perspective on Multi-Domain and Multi-Task Learning (2014, methodology): “The noises are synthesised by Audio Degradation Toolbox (Mauch & Ewert, 2013).”
A Software Framework for Musical Data Augmentation (2015, methodology): “Our implementation takes substantial inspiration from the Audio Degradation Toolbox [16].”
A Simple Method to Determine if a Music Information Retrieval System is a “Horse” (2014, methodology): “One might see the method of irrelevant transformations as robustness evaluation [27], which aims to answer questions of how the FoM of a system changes with environmental conditions.”