SPEECH RECOGNITION WITH DEEP LEARNING

  • Abel Gezahegn, Ajay Dureja, Aman Dureja
Keywords: Deep Learning, Speech Recognition, Convolutional neural network

Abstract

Speech recognition is the field of science involving the enablement of machines to accept voice information from humans and understanding it with the greatest likelihood of correctness. There are many methods to apply speech recognition models. One of the new approaches is utilizing neural networks with deep learning for speech recognition using a convolutional neural network(CNN). It is well recognized for its applicability to image recognition. Yet, another primary application of the technology is in speech recognition, such as Apple Siri, Microsoft Cortana, Google Assistant, Amazon Alexa. CNN eliminates spectral differences and predicts spectral associations found in the signals. This paper will collect online voice data, and noise elimination on the voice data and data should be filtered for more process Then, the filtered voice data is then passed into the CNN model and trained through different layers. After this, the trained model is checked for consistency, validity, accuracy, and then the trained data is checked with test data to validate the test accuracy of a given Model.

Published
2021-08-28
How to Cite
Aman Dureja, A. G. A. D. (2021). SPEECH RECOGNITION WITH DEEP LEARNING. Design Engineering, 4360- 4369. Retrieved from http://thedesignengineering.com/index.php/DE/article/view/3827
Section
Articles