Exploration of diverse intelligent approaches in speech recognition systems
- PDF / 938,563 Bytes
- 10 Pages / 595.276 x 790.866 pts Page_size
- 54 Downloads / 168 Views
Exploration of diverse intelligent approaches in speech recognition systems Iwin Thanakumar Joseph Swamidason1 · Sravanthi Tatiparthi2 · V. M. Arul Xavier1 · C. S. C. Devadass3 Received: 4 May 2020 / Accepted: 29 October 2020 © Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Artificial Intelligence revolutionizes the industrial sector to the greater extent towards the era of smart world. Real time automatic speech recognition system is on greater demand for the past few years in most of the embedded devices and smart phone applications. Research on automatic speech recognition is quite challenging due to the complication of environmental noises especially with the non stationary one. Machine learning based robust models are developed widely for speech recognition applications in the past decades. Now the researches mostly focused on deep learning approaches in order to improve the performance and better results. The complexity in designing separate feature extraction steps and classification models in the earlier models are eliminated in the deep learning models. This research article presents the detailed view of various research models developed for the application of automatic speech recognition, its advantages and also the various deep learning frame works for exploring future works. Keywords Deep learning · Speech recognition system · Neural network · Artificial intelligence
1 Introduction Automatic speech recognition research performance (Zhang et al. 2018; Kumar and Singh 2019) is improved a lot during the past few years due to its need in enormous human–machine communication systems like Amazon, echo, Xbox, cortana, google Now etc. These smart applications are highly essential and be part of day to day human activities. Still there are few setbacks in these applications that given below which deteriorates the performance of these systems. 1. Addition of noises 2. Reverberation in microphone
* Iwin Thanakumar Joseph Swamidason [email protected] 1
Department of Computer Science and Engineering, Karunya Institute of Technology and Sciences, Coimbatore, India
2
Department of Computer Science and Engineering, Vaagdevi Engineering College, Warangal, India
3
Department of Civil Engineering, Samskruti College of Engineering and Technology, Hyderabad, India
Based on the spectral distribution, there are two categories of noises 1. Stationary noise 2. Non-stationary noise The research is grown to the extent that the targeted signal absence is identified at the particular instant, tackling of additive noise with regard to the standards. There are lots of signal processing techniques developed in the year between 1970 and 1980 for the reduction of noises especially unsupervised one. Still the challenge for detection and suppression of the effects of non stationary noises exists. These limitations are addressed effectively by means of data driven approaches that rely on supervised machine learning models and it improved the robustness of automatic speech recognition s
Data Loading...
