Infrastructure

Cable - Wireless

Broadband corDECT

Broadband optiMA
CygNet NMS
Soft Switch
Access devices
802.11 Wireless access

Applications

Rural ATM

Medical diagnostic kit

Online Tutorial
Indic computing
Video-conference

Rural Services

Rural Crafts
Rural ITES
Agri Initiatives

 

 

 

 

 

Products - Indic Computing


Multi computer interfaces (Indic computing)

Although India's average literacy level is about 65%, less than 5% of India's population can use English for communication. And even though the world wide web and computer communication has given us access to information at the click of a mouse, 95% of our population is excluded from this revolution due to dominance of English.

To overcome this problem, a multimodal interface to the computer that is relevant for India, i.e., one that enables Indic computing has been developed.

The components of this Indian language interface are:

  1. Keyboard and display interface
  2. Speech interface
  3. Handwriting interface

The focus is to integrate these interfaces into the kernel/operating system so that any application that runs on top it inherits the interface.

We currently have a system in place that supports a number of different languages when the keyboard is used as a mechanism for data entry. The speech interface supports limited vocabulary speech recognition. The handwriting recogniser performs online handwriting recognition. It supports Tamil. The speech and handwriting interfaces have also been integrated into KWord, the opensource word processor. The handwriting interface has also been successfully interfaced to a browser.

The speech output interface uses speech synthesis as its core technology. Similar to its input interface counterpart, it will not replace the monitor but supplement it. Our goal is to synthesize natural sounding speech in all the official Indian languages. We have currently a speech synthesiser for Hindi, Tamil and Telugu. The speech synthesiser uses a multilingual database of diphone units as the basic units. The quality of synthesis is intelligible but not natural. We are now building a speech synthesiser based on syllable-like units for Indian languages.

Back to Top

 
Copyright 1996 - 2005 © TeNeT
The Telecommunication and Computer Networking Group

Department of Electrical Engineering + Computer Science & Engineering, IIT Madras, Chennai 600036, India.

e-mail: office@tenet.res.in   Website developed & maintained by NNSpiders