Baseline Automatic Speech Recognition System Research Articles

Lexical modeling for the development of Amharic automatic speech recognition systems

Amharic is the second most spoken Semitic language after Arabic. It has its own syllabary writing system, each character representing a consonant and a vowel. Automatic Speech Recognition (ASR) researches for Amharic have been conducted on the basis of grapheme-based pronunciation lexicon, taking advantage of the nature of its writing system. However, the epenthetic vowel and the glottal stop consonant represented in the writing system may not be pronounced in all of their occurrences. Moreover, the writing system does not differentiate geminated and non-geminated forms of consonants. Therefore, the grapheme-based pronunciation lexicon used so far has limitations with regard to these language features. To handle these limitations, we have prepared word- and morpheme-based pronunciation lexicons using data-driven and knowledge-driven experts’ transcription. The data-driven transcription has been used for the preparation of training pronunciation lexicon while the knowledge-driven has been used to prepare morpheme- and word-based pronunciation lexicons for decoding. When morpheme-based knowledge-driven lexicons are used, better ASR performance (compared with the baseline ASR system that used grapheme-based lexicon) has been achieved although the number of phones is much more (60) than the number of phones used in the grapheme-based lexicon (37).

Improving low-resource Tibetan end-to-end ASR by multilingual and multilevel unit modeling

Conventional automatic speech recognition (ASR) and emerging end-to-end (E2E) speech recognition have achieved promising results after being provided with sufficient resources. However, for low-resource language, the current ASR is still challenging. The Lhasa dialect is the most widespread Tibetan dialect and has a wealth of speakers and transcriptions. Hence, it is meaningful to apply the ASR technique to the Lhasa dialect for historical heritage protection and cultural exchange. Previous work on Tibetan speech recognition focused on selecting phone-level acoustic modeling units and incorporating tonal information but underestimated the influence of limited data. The purpose of this paper is to improve the speech recognition performance of the low-resource Lhasa dialect by adopting multilingual speech recognition technology on the E2E structure based on the transfer learning framework. Using transfer learning, we first establish a monolingual E2E ASR system for the Lhasa dialect with different source languages to initialize the ASR model to compare the positive effects of source languages on the Tibetan ASR model. We further propose a multilingual E2E ASR system by utilizing initialization strategies with different source languages and multilevel units, which is proposed for the first time. Our experiments show that the performance of the proposed method-based ASR system exceeds that of the E2E baseline ASR system. Our proposed method effectively models the low-resource Lhasa dialect and achieves a relative 14.2% performance improvement in character error rate (CER) compared to DNN-HMM systems. Moreover, from the best monolingual E2E model to the best multilingual E2E model of the Lhasa dialect, the system’s performance increased by 8.4% in CER.

Block‐online multi‐channel speech enhancement using deep neural network‐supported relative transfer function estimates

This work addresses the problem of block-online processing for multi-channel speech enhancement. Such processing is vital in scenarios with moving speakers and/or when very short utterances are processed, e.g., in voice assistant scenarios. We consider several variants of a system that performs beamforming supported by DNN-based voice activity detection (VAD) followed by post-filtering. The speaker is targeted through estimating relative transfer functions between microphones. Each block of the input signals is processed independently in order to make the method applicable in highly dynamic environments. Owing to the short length of the processed block, the statistics required by the beamformer are estimated less precisely. The influence of this inaccuracy is studied and compared to the processing regime when recordings are treated as one block (batch processing). The experimental evaluation of the proposed method is performed on large datasets of CHiME-4 and on another dataset featuring moving target speaker. The experiments are evaluated in terms of objective and perceptual criteria (such as signal-to-interference ratio (SIR) or perceptual evaluation of speech quality (PESQ), respectively). Moreover, word error rate (WER) achieved by a baseline automatic speech recognition system is evaluated, for which the enhancement method serves as a front-end solution. The results indicate that the proposed method is robust with respect to short length of the processed block. Significant improvements in terms of the criteria and WER are observed even for the block length of 250 ms.

Nonparametric Bayesian Double Articulation Analyzer for Direct Language Acquisition From Continuous Speech Signals

Human infants can discover words directly from unsegmented speech signals without any explicitly labeled data. In this paper, we develop a novel machine learning method called nonparametric Bayesian double articulation analyzer (NPB-DAA) that can directly acquire language and acoustic models from observed continuous speech signals. For this purpose, we propose an integrative generative model that combines a language model and an acoustic model into a single generative model called the "hierarchical Dirichlet process hidden language model" (HDP-HLM). The HDP-HLM is obtained by extending the hierarchical Dirichlet process hidden semi-Markov model (HDP-HSMM) proposed by Johnson et al. An inference procedure for the HDP-HLM is derived using the blocked Gibbs sampler originally proposed for the HDP-HSMM. This procedure enables the simultaneous and direct inference of language and acoustic models from continuous speech signals. Based on the HDP-HLM and its inference procedure, we developed a novel double articulation analyzer. By assuming HDP-HLM as a generative model of observed time series data, and by inferring latent variables of the model, the method can analyze latent double articulation structure, i.e., hierarchically organized latent words and phonemes, of the data in an unsupervised manner. The novel unsupervised double articulation analyzer is called NPB-DAA. The NPB-DAA can automatically estimate double articulation structure embedded in speech signals. We also carried out two evaluation experiments using synthetic data and actual human continuous speech signals representing Japanese vowel sequences. In the word acquisition and phoneme categorization tasks, the NPB-DAA outperformed a conventional double articulation analyzer (DAA) and baseline automatic speech recognition system whose acoustic model was trained in a supervised manner.

Comparing human and automatic speech recognition in a perceptual restoration experiment

Speech that has been distorted by introducing spectral or temporal gaps is still perceived as continuous and complete by human listeners, so long as the gaps are filled with additive noise of sufficient intensity. When such perceptual restoration occurs, the speech is also more intelligible compared to the case in which noise has not been added in the gaps. This observation has motivated so-called ‘missing data’ systems for automatic speech recognition (ASR), but there have been few attempts to determine whether such systems are a good model of perceptual restoration in human listeners. Accordingly, the current paper evaluates missing data ASR in a perceptual restoration task. We evaluated two systems that use a new approach to bounded marginalisation in the cepstral domain, and a bounded conditional mean imputation method. Both methods model available speech information as a clean-speech posterior distribution that is subsequently passed to an ASR system. The proposed missing data ASR systems were evaluated using distorted speech, in which spectro-temporal gaps were optionally filled with additive noise. Speech recognition performance of the proposed systems was compared against a baseline ASR system, and with human speech recognition performance on the same task. We conclude that missing data methods improve speech recognition performance in a manner that is consistent with perceptual restoration in human listeners.

Discriminative re-ranking for automatic speech recognition by leveraging invariant structures

An invariant structure was proposed in Minematsu (2004) and Minematsu et al. (2010) and it is a long-span feature to suppress non-linguistic factors. In contrast to frame-based features such as Mel-Frequency Cepstrum Coefficients (MFCC), the invariant structures are extracted as contrasts between speech events in a given utterance. Because the invariant structure is not a time series of short-term features, it is difficult to use it directly in the general framework of Automatic Speech Recognition (ASR) although its robustness against non-linguistic factors is desirable for ASR. To introduce the invariant structure effectively to ASR, we are working on a method to leverage the invariant structure in a discriminative re-ranking paradigm for ASR. In our re-ranking paradigm, a baseline ASR system is used to generate N-best lists with hypothesized phoneme-level alignments so that we can extract one invariant structure for each hypothesis. We also propose methods to convert an extracted invariant structure into a fixed-dimensional feature vector to be used in discriminative re-ranking. Experimental results on the three tasks of continuous digit recognition, digit recognition in noisy environments, and large vocabulary continuous speech recognition showed significant error reductions and robustness improvements against noisy environments.

Integration of Statistical Models for Dictation of Document Translations in a Machine-Aided Human Translation Task

This paper presents a model for machine-aided human translation (MAHT) that integrates source language text and target language acoustic information to produce the text translation of source language document. It is evaluated on a scenario where a human translator dictates a first draft target language translation of a source language document. Information obtained from the source language document, including translation probabilities derived from statistical machine translation (SMT) and named entity tags derived from named entity recognition (NER), is incorporated with acoustic phonetic information obtained from an automatic speech recognition (ASR) system. One advantage of the system combination used here is that words that are not included in the ASR vocabulary can be correctly decoded by the combined system. The MAHT model and system implementation is presented. It is shown that a relative decrease in word error rate of 29% can be obtained by this combined system relative to the baseline ASR performance on a French to English document translation task in the Hansard domain. In addition, it is shown that transcriptions obtained by using the combined system show a relative increase in NIST score of 34% compared to transcriptions obtained from the baseline ASR system.

A Hybrid Acoustic and Pronunciation Model Adaptation Approach for Non-native Speech Recognition

In this paper, we propose a hybrid model adaptation approach in which pronunciation and acoustic models are adapted by incorporating the pronunciation and acoustic variabilities of non-native speech in order to improve the performance of non-native automatic speech recognition (ASR). Specifically, the proposed hybrid model adaptation can be performed at either the state-tying or triphone-modeling level, depending at which acoustic model adaptation is performed. In both methods, we first analyze the pronunciation variant rules of non-native speakers and then classify each rule as either a pronunciation variant or an acoustic variant. The state-tying level hybrid method then adapts pronunciation models and acoustic models by accommodating the pronunciation variants in the pronunciation dictionary and by clustering the states of triphone acoustic models using the acoustic variants, respectively. On the other hand, the triphone-modeling level hybrid method initially adapts pronunciation models in the same way as in the state-tying level hybrid method; however, for the acoustic model adaptation, the triphone acoustic models are then re-estimated based on the adapted pronunciation models and the states of the re-estimated triphone acoustic models are clustered using the acoustic variants. From the Korean-spoken English speech recognition experiments, it is shown that ASR systems employing the state-tying and triphone-modeling level adaptation methods can relatively reduce the average word error rates (WERs) by 17.1% and 22.1% for non-native speech, respectively, when compared to a baseline ASR system.

Gaze-contingent automatic speech recognition

There has been progress in improving speech recognition using a tightly-coupled modality such as lip movement; and using additional input interfaces to improve recognition of commands in multimodal human–computer interfaces such as speech and pen-based systems. However, there has been little work that attempts to improve the recognition of spontaneous, conversational speech by adding information from a loosely–coupled modality. The study investigated this idea by integrating information from gaze into an automatic speech recognition (ASR) system. A probabilistic framework for multimodal recognition was formalised and applied to the specific case of integrating gaze and speech. Gaze-contingent ASR systems were developed from a baseline ASR system by redistributing language model probability mass according to the visual attention. These systems were tested on a corpus of matched eye movement and related spontaneous conversational British English speech segments (n=1355) for a visual-based, goal-driven task. The best performing systems had similar word error rates to the baseline ASR system and showed an increase in keyword spotting accuracy. The core values of this work may be useful for developing robust speech-centric multimodal decoding system functions.