By SHUBASHREE DESIKAN/ The Hindu

It is easy to think of applications for visually challenged people in the context of "text-to-speech" software (TTS), but now there is an application which uses TTS to help out farmers. The Sandesh Pathak application, developed jointly by C-DAC Mumbai, IIT-Madras, IIIT Hyderabad, IIT Kharagpur, and C-DAC Thiruvananthapuram will enable SMS messages to be read out loud, for the benefit of farmers who may have difficulty in reading.

"It is usable by people who cannot read. A large population of farmers belongs to this category. So when they receive an SMS message either containing agriculture-related advice or some other thing, this app will read aloud the content," says Pranaw Kumar, Senior Technical Officer, C-DAC, Mumbai who has been involved in developing this software.

The app which is available for download from the Appstore of the Mobile Seva Project of government of India, is an Indian language SMS Reader. It takes the received SMS message as input and reads it out aloud. It supports five Indian languages - Hindi, Tamil, Marathi, Gujarathi and Telugu. It also has options to select the language and speed of the voice as normal, slow, slower, fast and faster.

It uses the text-to-speech synthesis systems developed by the Indian Language TTS Consortium. To make it especially useful for farmers, "the TTS engines of all these languages have been tested on the agriculture domain-related texts and fine-tuned accordingly," he says.

The app is part of the project launched by the Indian Government to help farmers read messages which may be of the following types: advice to solve farming problems - insect, disease, fertilizer or weed management; information on weather - such as forecasts; and updates on latest technology - for improving yield and much more.

It is also possible that the text message may contain words in different languages, for example botanical names of plants or fertiliser names. Would the software be able to tackle this?

Pranaw Kumar says, "If the names have been transliterated in the language of the text it is possible for the app to read it out, but currently it is not equipped to read out messages which are bilingual. However, efforts are on to improve it to a version which has bilingual capacity."

In Phase two, where the plan is to include more languages, more institutes have been roped in - IIT Guwahati, IIT Mandi, SSN College Chennai, DA-IICT Ahmedabad, University of Guwahati and IISc, Bangalore. Work is on to make the app bilingual and to develop it in 13 languages, including Malayalam, Kannada, Rajasthani, Assamese, Manipuri, Odia and Bodo and also 13 flavours of English (with regional accents).

These Indian Language TTS systems have been developed under the Indian Language TTS Consortium which is led by IIT-Madras, under the leadership of Prof. Hema Murthy. C-DAC Mumbai is also a member of this collaboration.