An elitist approach for extracting automatically well-realized speech sounds with high confidence

TL;DR

This paper introduces an elitist method utilizing Hidden Markov Models to automatically extract high-confidence, well-realized speech sounds with improved detection reliability.

Contribution

It proposes a novel elitist approach that iteratively trains HMMs on well-detected speech sounds to enhance recognition accuracy.

Findings

High detection accuracy of specific speech sounds

Reduced error rates in speech recognition

Effective iterative training process

Abstract

This paper presents an "elitist approach" for extracting automatically well-realized speech sounds with high confidence. The elitist approach uses a speech recognition system based on Hidden Markov Models (HMM). The HMM are trained on speech sounds which are systematically well-detected in an iterative procedure. The results show that, by using the HMM models defined in the training phase, the speech recognizer detects reliably specific speech sounds with a small rate of errors.