Abstract: Natural Language Processing (NLP) is a field of software engineering, counterfeit consciousness and computational semantics which were worried with collaborations in the middle of PCs and human languages. Machine interpretation is a standout amongst the most essential uses of Natural Language Processing. This is the one of the vital technique that helps people from different places to understand an unknown language without the help of a human translator .The language to be translated is the Source Language (SL) and the language to which source language translated is Target Dialect (TL). While translating, the syntactic structure and semantics structure of both source and target language ought to be considered. Although there has been a ton of works distributed for widespread languages like English, works in language like Malayalam is similarly less. One of the most rising research area is in NLP and machine translation, which is from English to Malayalam. In this field this paper is centering upon the speech translation into Malayalam. It has an assortment of uses in the spaces like magazines,books,film reviews,etc.The approach included in this procedure is as follows. Here we propose a technique for the translation of English sentences to Malayalam and afterward into speech. This machine translation is done by rule based method and statistical method combined. The core process in this translation is mediated by bilingual dictionaries and rules for converting source language structures into the target language structures. The rules that are to be used in this approach are prepared based on the Parts Of Speech (POS) tags and dependency information that are obtained from the parser. There are mainly 2 types of rules are used here, one is transfer link rule and the other is morphological rules. In this method, for generating target structure, transfer link rules are used .Morphological rules used here are for assigning morphological features. The bilingual dictionary used here is an English, Malayalam bilingual dictionary. By this approach, a given English sentence can be translated to its Malayalam equivalent. And finally statistical method is used to reduce the error in the translation. Thus the first phase is completed, and in the next phase the translation of Malayalam text to speech is done, which is the main area on which this paper is focusing on. Hence it is concluded that speech translation is a very important area to be considered in Malayalam Language. Smart sermonis is a translator which convert English text to Malayalam speech. It provide a short but a comprehensive overview of Text-To-Speech synthesis by highlighting its digital signal processing component.
Keywords: Component, TTS, HMM Synthesis, Phoneme, Prosody, Concatenation Synthesis.