Hybrid machine translation from Marathi to English: A research evolution in machine translation

Abstract

India is multilingual country, in India people speak different dialect. Hence data present in different
dialect and structure gives rise to dialect as a barrier in information retrieval. In state like Maharashtra where
Marathi is the local dialect commonly used by the people, it gives rise to a non-adjustable environment for nonmarathi users to live in. Vital information on current happenings on village and taluka level published in
newspaper, destination and their distances indicated by the sign boards, advertisements and shops/hotels
names/other vital information presented with local language causes a major setback among Masses.
Information can be present in form of text, speech and image translating this information helps for sharing of
information and ultimately information gain.The proposed system uses Image Processing and Machine Aided
Translation for overcoming the above problem. Image processing is used because we are taking images as input
and extracting useful information from it which has to be translated. Machine aided Translation helps to
translate Information displayed in one dialect to other language(in our case from marathi to english). We are
presenting the output not only in text format but also in speech. Currently System has been proposed for images
which can be extended to speech and voice in future. Also many other dialects can be incorpated for the
betterment of the user.