Welcome to Francis Academic Press

The Frontiers of Society, Science and Technology, 2020, 2(11); doi: 10.25236/FSST.2020.021117.

Speech Recognition and Optimization Using Linear Classification Artificial Neural Network


Jingbo Cui1, Ting Liu2*,Xinkai Hao3*

Corresponding Author:
Ting Liu,Xinkai Hao

1 College of Information and Computer Science, Xi’an Jiaotong-Liverpool University, Suzhou 215123, China

2 College of electronic and Information Engineering, Beijing Jiaotong University, Beijing 100044, China

3 College of marine science and technology, Northwestern Polytechnical University, Xi'an 710072, China

*These authors contributed equally to this work and should be considered co-first authors.


This research studies the speech recognition process, and divides the speech recognition of linear system into four steps – speech acquisition, training, classification and results. For each part, its optimization is given. First, the effects of different feature sets of the same speech on classification results were tested. Then optimal parameter values of the neural network are found. Second, test the effect of different speech signal processing methods on speech recognition results. Present an analysis that shows whether STFT and ASTFT processing methods are effective in reducing error rate. Modify a neural network with four outputs to classify more digits. Third, the training step was modified from 10 outputs to 4 outputs (decimal to binary) and nCCs were transferred to binary for optimizing.


Neural network, Liner classification, Mscc, Stft

Cite This Paper

Jingbo Cui, Ting Liu,Xinkai Hao. Speech Recognition and Optimization Using Linear Classification Artificial Neural Network. The Frontiers of Society, Science and Technology (2020) Vol. 2 Issue 11: 117-126. https://doi.org/10.25236/FSST.2020.021117.


[1] Maruf A.Dhali, Camilo Nathan Jansen, Jan Willem de Wit, Lambertb Schomaker (2020). Feature-extraction methods for historical manuscript dating based on writing style development. Pattern Recognition Letters, vol.131, no3, pp.2020.

[2]  Nawel SOUISSI, Adnane CHERIF (2016). Speech Recognition System Based on Short-term Cepstral Parameters, Feature Reduction Method and Artificial Neural Networks. 2nd International Conference on Advanced Technologies for Signal and Image Processing ATSIP', no.3, pp.21-24.

[3] Anukul Anand, Manoj Kumar Mukul (2016). Comparison of STFT Based Direction of Arrival EstimationTechniques for Speech Signal. IEEE International Conference on Recent Trends in Electronics Information Communication Technology, no.5, pp.20-21.