Distinctive Phonetic Feature (DPF) Extraction Based on MLNs and Inhibition/Enhancement Network

Abstract

SUMMARYThis paper describes a distinctive phonetic feature (DPF) extraction method for use in a phoneme recognition system; our method has a low computation cost. This method comprises three stages. The first stage uses two multilayer neural networks (MLNs): MLN LF−DPF , which maps continuous acoustic features, or local features (LFs), onto discrete DPF features, and MLN Dyn , which constrains the DPF context at the phoneme boundaries. The second stage incorporates inhibition/enhancement (In/En) functionalities to discriminate whether the DPF dynamic patterns of trajectories are convex or concave, where convex patterns are enhanced and concave patterns are inhibited. The third stage decorrelates the DPF vectors using the Gram-Schmidt orthogonalization procedure before feeding them into a hidden Markov model (HMM)-based classifier. In an experiment on Japanese Newspaper Article Sentences (JNAS) utterances, the proposed feature extractor, which incorporates two MLNs and an In/En network, was found to provide a higher phoneme correct rate with fewer mixture components in the HMMs. key words: distinctive phonetic feature, hidden Markov model, multilayer neural network, inhibition/enhancement network, local features

Effects of Syllable Language Model on Distinctive Phonetic Features (DPFs) based Phoneme Recognition Performance

“…In this method, three MLNs instead of two MLNs [5] MLN LF-DPF , outputs DPFs [11,12] for the inputted acoustic features, LFs [15], while the second MLN, MLN cntxt , reduces misclassification at phoneme boundaries by taking seven frame context (from t-3 to t+3) as input, and the third MLN, MLN Dyn , restricts the DPF dynamics by incorporating dynamic parameters ( DPF and DPF) into its input. Here, the MLN LF-DPF , which is trained using the standard back-propagation learning algorithm, has two hidden layers of 256 and 96 units, respectively and takes three input vectors (t-3, t, t+3) of LFs of 25 dimensions each.…”

“…This phoneme misclassification sometimes occurs when the values of DPF peaks and DPF dips are closer to each other. Therefore, a mechanism, which is called In/En network [5,6,7], is needed to obtain clearly separable DPF peaks and dips. An algorithm for this network is given below:…”

“…Though various distinctive phonetic features (DPFs) based methods, which can solve coarticulatory phenomena [3,4], were proposed to obtain a more accurate phoneme recognizer [5,6,7,8], they provide a higher phoneme correct rate (PCR) with poor phoneme accuracy rate (PAR). The reason for providing lower phoneme recognition accuracy is the violation of some phonotactic constraints at different places of the output phoneme string of an input utterance.…”

Inhibition/Enhancement Network Based ASR using Multiple DPF Extractors

“…Various methods of phoneme recognition based on Inhibition/Enhancement (In/En) network were proposed by Huda, et al [1][2][3][4]. These papers introduced In/En functionality to discriminate whether the distinctive phonetic features (DPFs) dynamic patterns of trajectories are convex or concave.…”

“…These papers showed that In/En network has an effect of improving phoneme recognition performance in clean acoustic environment. The impact of In/En network in practical condition was not analyzed in [1][2][3][4].…”