Voice recognition in intelligent home systems

DOI №______

Abstract

In this article, an algorithm for voice recognition is presented for identifying a person based on the Gabor transformation and subsequent implementation of the system intelligent home voise messege. The idea of recognition is that guests and outsiders can not manage the system. The proposed approach is based on the creation of a spectrogramm that listens to a voice base and from which the features of the voice are distinguished through the heuristic algorithm, then the voice of the person is recognized, and then the command is executed and the answer is pronounced. A person's check is carried out with the use of the classical neural network. Represented on the spectrogram of the process of finding the voice of the host. From the drawings it is seen that when the number of sample samples increases, the recognition accuracy increases. A pre-created database of 20,25,75,100 voice samples has been tested in real time with various extraneous noises and sounds. Improvement of identification is carried out by the algorithm of voice recognition, the algorithm is taught in each voice command. With the help of the developed algorithm, it is possible to significantly improve the quality of perception of voice commands by identifying a person and removing noise from the signal. Theoretically, it is possible to say the command quietly and unclearly, but for this purpose it is necessary to develop scripts of the dialogue of the owner with the system. A large database of voice samples will greatly improve the quality. This algorithm is able to add real-time voice samples to the database, that is, when the host speaks the command, the algorithm selects the voice signs, then performs voice identification, then recognizes the command and executes it, at the end the program saves the voice sample by preliminarily filtering the outside sounds and noises.

Key words: neural network, voice recognition, person identification, Gabor transformation, heuristic algorithm, intelligent home.

References (MLA)
1. Awan S. N., Roy N., Zhang D., and Cohen S. M. " Validation of the Cepstral Spectral Index of Dysphonia as a Screening Tool for Voice Disorders: Development of Clinical Cutoff Scores." Journal of Voice 2(30) (2016): 130–144. Print.
2. Cpalka K., Zalasi´nski M., and Rutkowski L. "A New Algorithm for Identity Verification Based on the Analysis of a Handwritten Dynamic Signature." Applied soft computing (43) (2016): 47–56. Print.
3. Gregor K., Danihelka I., Graves A., Rezende D. J., and Wierstra D. Draw: A Recurrent Neural Network for Image Generation. Https://arxiv.org/abs/1502.04623. (2015). Web.
4. Grycuk R., Gabryel M., Nowicki R., and Scherer R. "Content-Based Image Retrieval Optimization by Differential Evolution." IEEE Congress on (2016): 86–93. Print.
5. Kim J., Oh K., Teoh A. B.-J., and Toh K.-A. "Finger-Knuckle-Print for Identity Verification Based on Difference Images. " IEEE 11th Conference on (ICIEA) (2016): 1073–1077. Print.
6. Mirjalili S. "Dragonfly Algorithm: A New Meta-Heuristic Optimization Technique for Solving Single-Objective, Discrete, and Multi-Objective Problems." Neural Computing and Applications 4(27) (2016): 1053-1073. Print.
7. Pal M. and Saha G. "On Robustness of Speech Based Biometric Systems Against Voice Conversion Attack." Applied Soft Computing (30) (2015): 214–228. Print.
8. Scherer M., Grycuk R., Gabryel M., and Voloshynovskiy S. "Image Descriptor Based on Edge Detection and Crawler Algorithm. " International Conference on Artificial Intelligence and Soft Computing (2016): 647–659. Print.
9. Usha M., Geetha Y., and Darshan Y. "Objective Identification of Prepubertal Female Singers and Non-Singers by Singing Power Ratio Using Matlab." Journal of Voice (2016). Print.
10. Werbos P. "Beyond Regression: New Tools for Prediction and Analysis in the Behavioral Sciences."(1974). Print.
11. Williams D. and Hinton G. "Learning Representations dy Backpropagating Errors." Nature (323) (1986): 533–536. Print.
12. Krasnova E., Bulgakova E., and Shchemelinin V. Performance Evaluation of AcousticSpectrographic Voice Identification Method in Native and Non-Native Speech. Moscow, 2016. Print.