Analysis of 2D Feature Spaces for Deep Learning-Based Speech Recognition

Korvel G, Treigys P, Tamulevicus G, Bernataviciene J, Kostek B. Analysis of 2D Feature Spaces for Deep Learning-Based Speech Recognition. J Audio Eng Soc. 2018;66(12):1072-1081. doi:10.17743/jaes.2018.0066

@article{Korvel2018_19880,
  author = {Korvel, Gražina and Treigys, Povilas and Tamulevicus, Gintautas and Bernataviciene, Jolita and Kostek, Bozena},
  title = {{Analysis of 2D Feature Spaces for Deep Learning-Based Speech Recognition}},
  journal = {Journal of the Audio Engineering Society},
  volume = {66},
  number = {12},
  pages = {1072--1081},
  year = {2018},
  month = dec,
  publisher = {Audio Engineering Society},
  doi = {10.17743/jaes.2018.0066},
  url = {https://doi.org/10.17743/jaes.2018.0066}
}

TY  - JOUR
TI  - Analysis of 2D Feature Spaces for Deep Learning-Based Speech Recognition
AU  - Korvel, Gražina
AU  - Treigys, Povilas
AU  - Tamulevicus, Gintautas
AU  - Bernataviciene, Jolita
AU  - Kostek, Bozena
T2  - Journal of the Audio Engineering Society
J2  - J. Audio Eng. Soc.
VL  - 66
IS  - 12
SP  - 1072
EP  - 1081
PY  - 2018
DA  - 2018/12/06
DO  - 10.17743/jaes.2018.0066
UR  - https://doi.org/10.17743/jaes.2018.0066
PB  - Audio Engineering Society
LA  - en
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.
ER  -