Extended VTS for noise-robust speech recognition

Model compensation is a standard way of improving the robustness of speech recognition systems to noise. A number of popular schemes are based on vector Taylor series (VTS) compensation, which uses a linear approximation to represent the influence of noise on the clean speech. To compensate the dynamic parameters, the continuous time approximation is often used. This approximation uses a point estimate of the gradient, which fails to take into account that dynamic coefficients are a function of a number of consecutive static coefficients. In this paper, the accuracy of dynamic parameter compensation is improved by representing the dynamic features as a linear transformation of a window of static features. A modified version of VTS compensation is applied to the distribution of the window of static features and, importantly, their correlations. These compensated distributions are then transformed to distributions over standard static and dynamic features. With this improved approximation, it is also possible to obtain full-covariance corrupted speech distributions. This addresses the correlation changes that occur in noise. The proposed scheme outperformed the standard VTS scheme by 10% to 20% relative on a range of tasks. © 2006 IEEE.

Scaling the iHMM: Parallelization versus Hadoop

This paper compares parallel and distributed implementations of an iterative, Gibbs sampling, machine learning algorithm. Distributed implementations run under Hadoop on facility computing clouds. The probabilistic model under study is the infinite HMM [1], in which parameters are learnt using an instance blocked Gibbs sampling, with a step consisting of a dynamic program. We apply this model to learn part-of-speech tags from newswire text in an unsupervised fashion. However our focus here is on runtime performance, as opposed to NLP-relevant scores, embodied by iteration duration, ease of development, deployment and debugging. © 2010 IEEE.

Bayesian adaptation and adaptively trained systems

As the use of found data increases, more systems are being built using adaptive training. Here transforms are used to represent unwanted acoustic variability, e.g. speaker and acoustic environment changes, allowing a canonical model that models only the "pure" variability of speech to be trained. Adaptive training may be described within a Bayesian framework. By using complexity control approaches to ensure robust parameter estimates, the standard point estimate adaptive training can be justified within this Bayesian framework. However during recognition there is usually no control over the amount of data available. It is therefore preferable to be able to use a full Bayesian approach to applying transforms during recognition rather than the standard point estimates. This paper discusses various approximations to Bayesian approaches including a new variational Bayes approximation. The application of these approaches to state-of-the-art adaptively trained systems using both CAT and MLLR transforms is then described and evaluated on a large vocabulary speech recognition task. © 2005 IEEE.

The infinite HMM for unsupervised PoS tagging

We extend previous work on fully unsupervised part-of-speech tagging. Using a non-parametric version of the HMM, called the infinite HMM (iHMM), we address the problem of choosing the number of hidden states in unsupervised Markov models for PoS tagging. We experiment with two non-parametric priors, the Dirichlet and Pitman-Yor processes, on the Wall Street Journal dataset using a parallelized implementation of an iHMM inference algorithm. We evaluate the results with a variety of clustering evaluation metrics and achieve equivalent or better performances than previously reported. Building on this promising result we evaluate the output of the unsupervised PoS tagger as a direct replacement for the output of a fully supervised PoS tagger for the task of shallow parsing and compare the two evaluations. © 2009 ACL and AFNLP.

Modeling natural sounds with modulation cascade processes

Natural sounds are structured on many time-scales. A typical segment of speech, for example, contains features that span four orders of magnitude: Sentences ($\sim1$s); phonemes ($\sim10$−$1$ s); glottal pulses ($\sim 10$−$2$s); and formants ($\sim 10$−$3$s). The auditory system uses information from each of these time-scales to solve complicated tasks such as auditory scene analysis [1]. One route toward understanding how auditory processing accomplishes this analysis is to build neuroscience-inspired algorithms which solve similar tasks and to compare the properties of these algorithms with properties of auditory processing. There is however a discord: Current machine-audition algorithms largely concentrate on the shorter time-scale structures in sounds, and the longer structures are ignored. The reason for this is two-fold. Firstly, it is a difficult technical problem to construct an algorithm that utilises both sorts of information. Secondly, it is computationally demanding to simultaneously process data both at high resolution (to extract short temporal information) and for long duration (to extract long temporal information). The contribution of this work is to develop a new statistical model for natural sounds that captures structure across a wide range of time-scales, and to provide efficient learning and inference algorithms. We demonstrate the success of this approach on a missing data task.

Probabilistic Amplitude Demodulation

Auditory scene analysis is extremely challenging. One approach, perhaps that adopted by the brain, is to shape useful representations of sounds on prior knowledge about their statistical structure. For example, sounds with harmonic sections are common and so time-frequency representations are efficient. Most current representations concentrate on the shorter components. Here, we propose representations for structures on longer time-scales, like the phonemes and sentences of speech. We decompose a sound into a product of processes, each with its own characteristic time-scale. This demodulation cascade relates to classical amplitude demodulation, but traditional algorithms fail to realise the representation fully. A new approach, probabilistic amplitude demodulation, is shown to out-perform the established methods, and to easily extend to representation of a full demodulation cascade.

Instantaneous frequency and bandwidth estimation using filterbank arrays

Accurate estimation of the instantaneous frequency of speech resonances is a hard problem mainly due to phase discontinuities in the speech signal associated with excitation instants. We review a variety of approaches for enhanced frequency and bandwidth estimation in the time-domain and propose a new cognitively motivated approach using filterbank arrays. We show that by filtering speech resonances using filters of different center frequency, bandwidth and shape, the ambiguity in instantaneous frequency estimation associated with amplitude envelope minima and phase discontinuities can be significantly reduced. The novel estimators are shown to perform well on synthetic speech signals with frequency and bandwidth micro-modulations (i.e., modulations within a pitch period), as well as on real speech signals. Filterbank arrays, when applied to frequency and bandwidth modulation index estimation, are shown to reduce the estimation error variance by 85% and 70% respectively. © 2013 IEEE.