101. Real Time Implementation of Text to Speech Synthesis
- Author
-
Sharma, Sushil Kumar and Sharma, Sushil Kumar
- Abstract
The purpose of a Text to Speech (TTS/T2S) synthesis is to provide artificial voice for a people and machine. Here a user can give input text using keypad and can get a similar synthesize voice using a speaker. This speech synthesis process involves first natural language processing and then digital signal processing to convert input text into processed text into speech. Many current text to speech synthesis is based on the concatenation of acoustic segments of prerecorded audio segments, and this speech synthesizer is capable of producing natural and intelligible speech, whose quality is very similar to the voice quality of the speaker who recorded the audio segments from which the concatenation process take places. Here, we proposed a real time functioned text to speech synthesis using STM32F4 discovery kit and MATLAB, which is capable to converts input text into speech and reads out to the user which can then be saved as a wave (.wav) file and/or text (.txt) file. Here a speech impairment person can type the text from keypad and using MATLAB R2015a, we can get voice output of input words, and output voice is heard through speaker, this will be of great help to visual impairment people to read and listen the large volume of text easier. The proposed methodology will be also great use of providing assistance to non-native people who lack the power of speech and many more applications in the field of multimedia and tele communication. In this thesis on real time implementation of text to speech, we presented design mythology and an algorithm to implement real time functioned text to speech synthesis for synthesizing unlimited vocabulary speech in for English language and all other languages scripted in Roman using STM32F4 Discovery kit (STM32F407VGT6 MCU) & MATLAB R2015a. The recording of voice is done in C language using STM32F4 Discovery kit and a database of acoustic library is constructed, whereas implementation of text to speech synthesis is done using MATLAB R2