Speaker-independent isolated word recognition using dynamic features of speech spectrum

TLDR

The study proposes a speaker‑independent isolated word recognition method that combines instantaneous and dynamic speech‑spectrum features. The method represents speech by time‑sequences of cepstrum coefficients and energy, extracts per‑frame regression coefficients (~50 ms), combines them with the original cepstrum, and uses a staggered‑array dynamic‑programming matcher to compare templates. Experiments on 100 Japanese city names achieved a 2.4 % error rate, outperforming the 6.2 % error rate obtained using only cepstrum coefficients.

Abstract

This paper proposes a new isolated word recognition technique based on a combination of instantaneous and dynamic features of the speech spectrum. This technique is shown to be highly effective in speaker-independent speech recognition. Spoken utterances are represented by time sequences of cepstrum coefficients and energy. Regression coefficients for these time functions are extracted for every frame over an approximately 50 ms period. Time functions of regression coefficients extracted for cepstrum and energy are combined with time functions of the original cepstrum coefficients, and used with a staggered array DP matching algorithm to compare multiple templates and input speech. Speaker-independent isolated word recognition experiments using a vocabulary of 100 Japanese city names indicate that a recognition error rate of 2.4 percent can be obtained with this method. Using only the original cepstrum coefficients the error rate is 6.2 percent.