On Variable-Scale Piecewise Stationary Spectral Analysis of Speech Signals for ASR

Hervé Bourlard, Vivek Tyagi
IDIAP, 2005
Report or working paper

Abstract

It is often acknowledged that speech signals contain short-term and long-term temporal properties that are difficult to capture and model by using the usual fixed scale (typically 20ms) short time spectral analysis used in hidden Markov models (HMMs), based on piecewise stationarity and state conditional independence assumptions of acoustic vectors. For example, vowels are typically quasi-stationary over 40-80ms segments, while plosives typically require analysis below 20ms segments. Thus, fixed scale analysis is clearly sub-optimal for ``optimal'' time-frequency resolution and modeling of different stationary phones found in the speech signal. In the present paper, we investigate the potential advantages of using variable size analysis windows towards improving state-of-the-art speech recognition systems. Based on the usual assumption that the speech signal can be modeled by a varying autoregressive (AR) Gaussian process, we estimate the largest piecewise quasi-stationary speech segments, based on the likelihood that a segment was generated by the same AR process. This likelihood is estimated from the Linear Prediction (LP) residual error. Each of these quasi-stationary segments is then used as an analysis window from which spectral features are extracted. Such an approach thus results in variable scale time spectral analysis, adaptively estimating the largest possible analysis window size such that the signal remains quasi-stationary, thus the best temporal/frequency resolution tradeoff. Speech recognition experiments on the OGI Numbers95 database show that the proposed multi-scale piecewise stationary spectral analysis based features indeed yield improved recognition accuracy in clean conditions, compared to features based on minimum cross entropy spectrum as well as those based on fixed scale spectral analysis.

Official source

https://infoscience.epfl.ch/record/83279?ln=en

About this result

This page is automatically generated and may contain information that is not correct, complete, up-to-date, or relevant to your search query. The same applies to every other page on this website. Please make sure to verify the information with EPFL's official sources.

Related concepts

Related publications

Hervé Bourlard, Vivek Tyagi
IDIAP, 2005
Report or working paper

Abstract

Official source

https://infoscience.epfl.ch/record/83279?ln=en

About this result

Related concepts (10)

Related concepts

Related publications (21)

Related publications

On Variable-Scale Piecewise Stationary Spectral Analysis of Speech Signals for ASR

Hervé Bourlard, Vivek Tyagi

It is often acknowledged that speech signals contain short-term and long-term temporal properties that are difficult to capture and model by using the usual fixed scale (typically 20ms) short time spectral analysis used in hidden Markov models (HMMs), based on piecewise stationarity and state conditional independence assumptions of acoustic vectors. For example, vowels are typically quasi-stationary over 40-80ms segments, while plosive typically require analysis below 20ms segments. Thus, fixed scale analysis is clearly sub-optimal for ``optimal'' time-frequency resolution and modeling of different stationary phones found in the speech signal. In the present paper, we investigate the potential advantages of using variable size analysis windows towards improving state-of-the-art speech recognition systems. Based on the usual assumption that the speech signal can be modeled by a time-varying autoregressive (AR) Gaussian process, we estimate the largest piecewise quasi-stationary speech segments, based on the likelihood that a segment was generated by the same AR process. This likelihood is estimated from the Linear Prediction (LP) residual error. Each of these quasi-stationary segments is then used as an analysis window from which spectral features are extracted. Such an approach thus results in a variable scale time spectral analysis, adaptively estimating the largest possible analysis window size such that the signal remains quasi-stationary, thus the best temporal/frequency resolution tradeoff. The speech recognition experiments on the OGI Numbers95 database, show that the proposed variable-scale piecewise stationary spectral analysis based features indeed yield improved recognition accuracy in clean conditions, compared to features based on minimum cross entropy spectrum as well as those based on fixed scale spectral analysis.

Elsevier2006

State-of-the-art automatic speech recognition (ASR) techniques are typically based on hidden Markov models (HMMs) for the modeling of temporal sequences of feature vectors extracted from the speech signal. At the level of each HMM state, Gaussian mixture models (GMMs) or artificial neural networks (ANNs) are commonly used in order to model the state emission probabilities. However, both GMMs and ANNs are rather rigid, as they are incapable of adapting to variations inherent in the speech signal, such as inter- and intra-speaker variations. Moreover, performance degradations of these systems are severe in the case of unmatched conditions such as in the presence of environmental noise. A lot of research effort is currently being devoted to overcoming these problems. The principal objective of this thesis is to explore new approaches towards a more robust and adaptive modeling of speech. In this context, different aspects of the modeling of speech data with HMMs and GMMs are investigated. Particular attention is given to the modeling of correlation. While correlation between different feature vectors (corresponding to temporal correlation) is typically modeled by the HMM, correlation between feature vector components (e.g., correlation in frequency) is modeled by the GMM part of the model. This thesis starts with the investigation of two potential ways to improve the modeling of correlation, consisting of (1) a shift of the modeling of temporal correlation towards GMMs, and (2) the modeling of correlation within each feature vector by a particular type of HMM. This leads to the development of a novel approach, referred to as ÒHMM2Ó, which is a major focus of this thesis. HMM2 is a particular mixture of hidden Markov models, where state emission probabilities of the temporal (primary) HMM are modeled through (secondary) state-dependent frequency-based HMMs. Low-dimensional GMMs are used for modeling the state emission probabilities of the secondary HMM states. Therefore, HMM2 can be seen as a generalization of conventional HMMs, which they include as a particular case. HMM2 may have several advantages as compared to standard systems. While the primary HMM performs time warping and time integration, the secondary HMM performs warping and integration along the frequency dimension of the speech signal. Frequency correlation is modeled through the secondary HMM topology. Due to the implicit, non-linear, state-dependent spectral warping performed by the secondary HMM, HMM2 may be viewed as a dynamic extension of the multi-band approach. Moreover, this frequency warping property may result in a better, more flexible modeling and parameter sharing. After an investigation of theoretical and practical aspects of HMM2, encouraging recognition results for the case of speech degraded by additive noise are given. Due to the spectral warping property of HMM2, this model is able to extract pertinent structural information of the speech signal, which is reflected in the trained model parameters. Consequently, such an HMM2 system can also be used to explicitly extract structures of a speech signal, which can then be converted into a new kind of ASR features, referred to as ÒHMM2 featuresÓ. In fact, frequency bands with similar characteristics are supposed to be emitted by the same secondary HMM state. The warping along the frequency dimension of speech thus results in an adaptable, data-driven frequency segmentation. In fact, as it can be assumed that different secondary HMM states model spectral regions characterized by high and low energies respectively, this segmentation may be related to formant structures. The application of HMM2 as a feature extractor is investigated, and it is shown that a system combining HMM2 features with conventional noise-robust features yields an improved speech recognition robustness. Moreover, a comparison of HMM2 features with formant tracks shows a comparable performance on a vowel classification task.

IDIAP2003

Automatic processing of multiparty interactions is a research domain with important applications in content browsing, summarization and information retrieval. In recent years, several works have been devoted to find regular patterns which speakers exhibit in a multiparty interaction also known as social roles. Most of the research in literature has generally focused on recognition of scenario specific formal roles. More recently, role coding schemes based on informal social roles have been proposed in literature, defining roles based on the behavior speakers have in the functioning of a small group interaction. Informal social roles represent a flexible classification scheme that can generalize across different scenarios of multiparty interaction. In this thesis, we focus on automatic recognition of informal social roles and exploit the influence of informal social roles on speaker behavior for structuring multiparty interactions. To model speaker behavior, we systematically explore various verbal and non verbal cues extracted from turn taking patterns, vocal expression and linguistic style. The influence of social roles on the behavior cues exhibited by a speaker is modeled using a discriminative approach based on conditional random fields. Experiments performed on several hours of meeting data reveal that classification using conditional random fields improves the role recognition performance. We demonstrate the effectiveness of our approach by evaluating it on previously unseen scenarios of multiparty interaction. Furthermore, we also consider whether formal roles and informal roles can be automatically predicted by the same verbal and nonverbal features. We exploit the influence of social roles on turn taking patterns to improve speaker diarization under distant microphone condition. Our work extends the Hidden Markov model (HMM)- Gaussian mixture model (GMM) speaker diarization system, and is based on jointly estimating both the speaker segmentation and social roles in an audio recording. We modify the minimum duration constraint in HMM-GMM diarization system by using role information to model the expected duration of speaker's turn. We also use social role n-grams as prior information to model speaker interaction patterns. Finally, we demonstrate the application of social roles for the problem of topic segmentation in meetings. We exploit our findings that social roles can dynamically change in conversations and use this information to predict topic changes in meetings. We also present an unsupervised method for topic segmentation which combines social roles and lexical cohesion. Experimental results show that social roles improve performance of both speaker diarization and topic segmentation.

EPFL2015

IDIAP2003

EPFL2015