A Method to offer multilingual TTS in embedded systems utilizing English/Persian as an example

Abstract

Most multilingual Test to Speech (TTS) systems are software applications which allow people with visual impairments or reading disabilities to listen to written material using computer. This paper describes an approach to make a multilingual TTS and embed it into the portable, low cost, and stand alone embedded system to access and read electronic documents particularly in developing countries. There are several TTS such as Doubletalk, DECtalk, and Dolphin available in market, also there are some products using TTS such as Talking OCR, Bill Reader and Intel Reader which are not affordable or multilingual. To design this system OMAP3530 an application processor board is considered as the hardware platform to process the language-independent parts of the application and RC8660 used as an integrated TTS processor.