Large Vocabulary Speech Recognition: Speaker Dependent and Speaker Independent

This paper addresses the problem of large vocabulary isolated word and continuous Kannada speech recognition using the syllables and combination of Hidden Markov Model (HMM) and Normal fit method. The models designed for speaker dependent and speaker independent mode of working. This experiment has covered 6 million words among the 10 million words from Hampi text corpus. Here 3-state Baum-Welch algorithm is used for training. For the 2 successor outputted lambda(A, B, pi) is combined and passed into normal fit, the outputted normal fit parameter is labeled has syllable or sub-word. In terms of memory requirement and recognition rate the proposed model is compared with Gaussian Mixture Model and HMM (3-state Baum-Welch algorithm). This paper clearly shows that combination of HMM and normal fit technique will reduce the memory size while building and storing the speech models and works with excellent recognition rate. The average WRR is 91.22 % and average WER is 8.78 %. All computations are done using mat lab.