AES E-Library

G. Korvel, P. Treigys, G. Tamulevicus, J. Bernataviciene, and B. Kostek, "Analysis of 2D Feature Spaces for Deep Learning-Based Speech Recognition," J. Audio Eng. Soc., vol. 66, no. 12, pp. 1072-1081, (2018 December.). doi: https://doi.org/10.17743/jaes.2018.0066 G. Korvel, P. Treigys, G. Tamulevicus, J. Bernataviciene, and B. Kostek, "Analysis of 2D Feature Spaces for Deep Learning-Based Speech Recognition," J. Audio Eng. Soc., vol. 66 Issue 12 pp. 1072-1081, (2018 December.). doi: https://doi.org/10.17743/jaes.2018.0066
Abstract: The aim of this study was to evaluate the suitability of 2D audio signal feature maps for speech recognition based on deep learning. The proposed methodology employs a convolutional neural network (CNN), which is a class of deep, feed-forward artificial neural network. The authors analyzed the audio signal feature maps, namely spectrograms, linear and Mel-scale cepstrograms, and chromagrams. This choice was made because CNN performs well in 2D data-oriented processing contexts. Feature maps were employed in a Lithuanian word-recognition task. The spectral analysis led to the highest word recognition rate. Spectral and mel-scale cepstral feature spaces outperform linear cepstra and chroma. The 111-word classification experiment depicts f1 score of 0.99 for spectrum, 0.91 for mel-scale cepstrum , 0.76 for chromagram, and 0.64 for cepstrum feature space on test data set.

@article{korvel2018analysis,
author={korvel, gražina and treigys, povilas and tamulevicus, gintautas and bernataviciene, jolita and kostek, bozena},
journal={journal of the audio engineering society},
title={analysis of 2d feature spaces for deep learning-based speech recognition},
year={2018},
volume={66},
number={12},
pages={1072-1081},
doi={https://doi.org/10.17743/jaes.2018.0066},
month={december},} @article{korvel2018analysis,
author={korvel, gražina and treigys, povilas and tamulevicus, gintautas and bernataviciene, jolita and kostek, bozena},
journal={journal of the audio engineering society},
title={analysis of 2d feature spaces for deep learning-based speech recognition},
year={2018},
volume={66},
number={12},
pages={1072-1081},
doi={https://doi.org/10.17743/jaes.2018.0066},
month={december},
abstract={the aim of this study was to evaluate the suitability of 2d audio signal feature maps for speech recognition based on deep learning. the proposed methodology employs a convolutional neural network (cnn), which is a class of deep, feed-forward artificial neural network. the authors analyzed the audio signal feature maps, namely spectrograms, linear and mel-scale cepstrograms, and chromagrams. this choice was made because cnn performs well in 2d data-oriented processing contexts. feature maps were employed in a lithuanian word-recognition task. the spectral analysis led to the highest word recognition rate. spectral and mel-scale cepstral feature spaces outperform linear cepstra and chroma. the 111-word classification experiment depicts f1 score of 0.99 for spectrum, 0.91 for mel-scale cepstrum , 0.76 for chromagram, and 0.64 for cepstrum feature space on test data set.},}

TY - report
TI - Analysis of 2D Feature Spaces for Deep Learning-Based Speech Recognition
SP - 1072 EP - 1081
AU - Korvel, Gražina
AU - Treigys, Povilas
AU - Tamulevicus, Gintautas
AU - Bernataviciene, Jolita
AU - Kostek, Bozena
PY - 2018
JO - Journal of the Audio Engineering Society
IS - 12
VO - 66
VL - 66
Y1 - December 2018 TY - report
TI - Analysis of 2D Feature Spaces for Deep Learning-Based Speech Recognition
SP - 1072 EP - 1081
AU - Korvel, Gražina
AU - Treigys, Povilas
AU - Tamulevicus, Gintautas
AU - Bernataviciene, Jolita
AU - Kostek, Bozena
PY - 2018
JO - Journal of the Audio Engineering Society
IS - 12
VO - 66
VL - 66
Y1 - December 2018
AB - The aim of this study was to evaluate the suitability of 2D audio signal feature maps for speech recognition based on deep learning. The proposed methodology employs a convolutional neural network (CNN), which is a class of deep, feed-forward artificial neural network. The authors analyzed the audio signal feature maps, namely spectrograms, linear and Mel-scale cepstrograms, and chromagrams. This choice was made because CNN performs well in 2D data-oriented processing contexts. Feature maps were employed in a Lithuanian word-recognition task. The spectral analysis led to the highest word recognition rate. Spectral and mel-scale cepstral feature spaces outperform linear cepstra and chroma. The 111-word classification experiment depicts f1 score of 0.99 for spectrum, 0.91 for mel-scale cepstrum , 0.76 for chromagram, and 0.64 for cepstrum feature space on test data set.