Data Augmentation using Reverb and Noise in Deep Learning Implementation of Cough Classification

Huq, S.; Xi, Pengcheng; Goubran, Rafik; Valdés, Julio J.; Knoefel, Frank; Green, James R.

doi:10.1109/memea57477.2023.10171862

Cited by 8 publications

(2 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This includes a short discussion on the dataset used, the augmentation methods evaluated, and the comparison of the performance of these methods. Portions of this chapter were previously published in proceedings of the 2023 Institute of Electrical and Electronics Engineers (IEEE) Medical Measurements and Applications (MeMeA) conference on which I was the primary author [33]. The chapter ends with detailed work on a pretrained model on cough classification on a separate database of cough sounds (COUGHVID [34]).…”

Section: Thesis Outlinementioning

confidence: 99%

See 1 more Smart Citation

Differentiation of Dry and Wet Cough Sounds using A Deep Learning Model and Data Augmentation

Huq

View full text Add to dashboard Cite

Automated classification of cough sounds has increased in importance due to factors such as the worldwide COVID-19 pandemic. To train such classification models requires a large dataset of cough sounds; however, it remains challenging to find sufficient expert-labelled training data. This thesis explores a novel form of audio data augmentation, where training cough sounds are corrupted with varying levels of reverberation and Gaussian noise. The combination of noise and reverberation is more effective than both traditional image-based augmentation techniques and either noise or reverberation alone, leading to near-human accuracy on a wet vs. dry cough classification task using a ResNet18 model across two cough datasets. Alignment between the training and testing environments is examined using the Speech-Commands audio dataset. While models trained with the closest reverb and noise level to the test environment gave the best results, the proposed audio augmentation technique produces models with robust performance across test environments.Finally, I would like to thank my family -your unconditional love and encouragement has been the fuel that powered me through this experience. To my wife, Fariha, thank you for your love and patience, especially the last months of the program. To my sister and family, appi (Rumpa), cb (Tinku), Rehan & Faraz, thank you for always being there to cheer me up and encourage me every time I needed it. To my mother and father, thank you for your love, kindness, patience, and your belief in me.v Table of ContentsCopyright .

show abstract

Section: Thesis Outlinementioning

confidence: 99%

“…It refers to the persistence of sound in an enclosed space after the sound source has stopped emitting [49]. Understanding and managing the effects of reverberation are crucial for accurate and reliable audio analysis [33].…”

Section: Noise and Reverberationmentioning

confidence: 99%