Speech recognition is the process of converting a speech signal to a sequence of words, by means of algorithms implemented as a computer program. This project analyzes the existing models for recognition of Nepali speech. Upon finding the shortcomings of the existing models, we move on to Ear Model, which is the simulation of how human ear and brain work together for speech recognition. This model eventually was found out to provide better accu- racy than conventional methods.
To further alleviate accuracy, provision for Nepali Dictionary checking and the database of syllables’ frequencies from Nepali corpus has also been devised. By using these techniques we have been able to achieve a speech recognition model that can be helpful to all seeking better model of speech recognition of Nepali language and other languages as well.