Comparative Analysis of Spoken Telugu digits using MFCC and LPCC via Hidden Markov Models

The problem of speech recognition deals with interpreting the message conveyed in a spoken word or sentence. It has much significance in various automated tasks such as controlling home appliances, mobile applications, robotic operations, etc. As the domain of Internet of Things (IoT) evolves, there is much scope to assist people with special needs in various ways in their day today routines. The main stakeholders for this type of applications include senior citizens, specially-abled people, and illiterates of different age groups. In IoT environment, speechenabled supportive systems has got a great significance with various challenges. The systems that deal with spoken languages with a tiny data sets for a specific problem wm be a challenging task. This paper studies spoken digit recognition problem using Hidden Markov Models with three different features. The paper discusses the methods and obtained results.