A Literature Review on Speech Recognition and Cyber Security Systems

Abstract

Abstract: People use communication to convey their thoughts, ideas, and feelings, as well as to understand the feelings of others. It is a dynamic, methodical, adaptable, commercial and continuous process. Visual, text, spoken, written, gestural, and other modalities of communication are among them. These types can frequently be converted into other formats. When a person cannot understand or recognize the type of communication being used, this step becomes extremely crucial. The communication gap between the specially abled i.e. blind, deaf, mute or differently abled i.e. people with disorders like dyslexia, arthritis etc. and the rest of the world can be closed by converting to a form that the rest of the world understands. The paper that follows focuses mostly on Speech to Text conversion using Deep Learning Algorithms like LPC, MFCC, PLP which provides an innovative, real-time, natural, and user-friendly means of interacting with a computer that is more familiar to humans. The converted speech data can be made full duplex by transmitting the data to the cloud. Hence the converted speech data has to be sent to the cloud over a wireless channel. Thus, all converted data can be posted to the cloud over a wireless channel and retrieved by network clients and users in the network can retrieve them using a web application designed specifically for this purpose. Client in the network communicate utilizing the internet and website applications. Hence, they need to be secure and protected. This has necessitated the development of a robust data concealing method. To secure the messages generated by this system, it employs a dynamic approach to cryptography. The following paper provides an overview of the various methodologies mentioned. Keywords: Speech Recognition Algorithms (LPC, MFCC, PLP), Wireless Communication, Encryption, Decryption, ESP8266, Text To Speech, Cloud Computing.