Toward end-to-end interpretable convolutional neural networks for waveform signals

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Vu, Linh, Tran, Thu, Lim, Wern-Han, Phan, Raphael
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913339975663616
author Vu, Linh
Tran, Thu
Lim, Wern-Han
Phan, Raphael
author_facet Vu, Linh
Tran, Thu
Lim, Wern-Han
Phan, Raphael
contents This paper introduces a novel convolutional neural networks (CNN) framework tailored for end-to-end audio deep learning models, presenting advancements in efficiency and explainability. By benchmarking experiments on three standard speech emotion recognition datasets with five-fold cross-validation, our framework outperforms Mel spectrogram features by up to seven percent. It can potentially replace the Mel-Frequency Cepstral Coefficients (MFCC) while remaining lightweight. Furthermore, we demonstrate the efficiency and interpretability of the front-end layer using the PhysioNet Heart Sound Database, illustrating its ability to handle and capture intricate long waveform patterns. Our contributions offer a portable solution for building efficient and interpretable models for raw waveform data.
format Preprint
id arxiv_https___arxiv_org_abs_2405_01815
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Toward end-to-end interpretable convolutional neural networks for waveform signals
Vu, Linh
Tran, Thu
Lim, Wern-Han
Phan, Raphael
Sound
Artificial Intelligence
Audio and Speech Processing
This paper introduces a novel convolutional neural networks (CNN) framework tailored for end-to-end audio deep learning models, presenting advancements in efficiency and explainability. By benchmarking experiments on three standard speech emotion recognition datasets with five-fold cross-validation, our framework outperforms Mel spectrogram features by up to seven percent. It can potentially replace the Mel-Frequency Cepstral Coefficients (MFCC) while remaining lightweight. Furthermore, we demonstrate the efficiency and interpretability of the front-end layer using the PhysioNet Heart Sound Database, illustrating its ability to handle and capture intricate long waveform patterns. Our contributions offer a portable solution for building efficient and interpretable models for raw waveform data.
title Toward end-to-end interpretable convolutional neural networks for waveform signals
topic Sound
Artificial Intelligence
Audio and Speech Processing
url https://arxiv.org/abs/2405.01815