Research Projects

At the time of disasters, a huge volume of distress calls should be handled and this calls for an unexpected surge in the human resource. Automation of such dialog through conversational artificial intelligence is the need of the hour. Most of the distress calls would be in the native language and there is a lot of scope to develop technology that can cater to this need. Towards this end, this project aims in the development of a spoken dialog system in Kannada language to manage calls at the time of disasters. The project has three major tasks namely, automatic speech recognition, text information extraction and parsing and the text to speech synthesis. All these tasks require exhaustive experimentation and analysis to build a conversational dialog agent. Machine learning techniques such as transformer and attention learning based architectures will be developed and deployed for the same cause.