Speech Based Command and Control System

In recent years, due to vast development of technology, speech interface is being implemented for machines i.e. human can communicate with machine by talking to it. Automatic speech recognition (ASR) on a machine requires conversion of spoken dialogues into text. Spoken interaction offers a solution when our hands and eyes are busy in some other work. Smart phones are widely being used not only for making call but also for clicking photographs, sending SMS, playing music, gaming, accessing internet, as tablet PC etc. This lets the speech technology to emerge and evolve for mobile phones. Technology that is used for developing and modeling desktop applications cannot be directly applied to mobile phone applications. There are various challenges that need to be catered while developing model for mobile phones. Issues like hardware support, memory optimization and storage can pose a grave threat and need special attention. The goal of this work is to develop efficient acoustic model, language model and pronunciation dictionary for mobile phones.