The effect of dialect mismatch on likelihood-maximising speech enhancement for noise-robust speech recognition

Traditional speech enhancement methods optimise signal-level criteria such as signal-to-noise ratio, but these approaches are sub-optimal for noise-robust speech recognition. Likelihood-maximising (LIMA) frameworks are an alternative that optimise parameters of enhancement algorithms based on state sequences generated for utterances with known transcriptions. Previous reports of LIMA frameworks have shown significant promise for improving speech recognition accuracies under additive background noise for a range of speech enhancement techniques. In this paper we discuss the drawbacks of the LIMA approach when multiple layers of acoustic mismatch are present – namely background noise and speaker accent. Experimentation using LIMA-based Mel-filterbank noise subtraction on American and Australian English in-car speech databases supports this discussion, demonstrating that inferior speech recognition performance occurs when a second layer of mismatch is seen during evaluation.

The application of phonetic distribution normalisation to likelihood-maximising speech enhancement for robust ASR

Traditional speech enhancement methods optimise signal-level criteria such as signal-to-noise ratio, but such approaches are sub-optimal for noise-robust speech recognition. Likelihood-maximising (LIMA) frameworks on the other hand, optimise the parameters of speech enhancement algorithms based on state sequences generated by a speech recogniser for utterances of known transcriptions. Previous applications of LIMA frameworks have generated a set of global enhancement parameters for all model states without taking in account the distribution of model occurrence, making optimisation susceptible to favouring frequently occurring models, in particular silence. In this paper, we demonstrate the existence of highly disproportionate phonetic distributions on two corpora with distinct speech tasks, and propose to normalise the influence of each phone based on a priori occurrence probabilities. Likelihood analysis and speech recognition experiments verify this approach for improving ASR performance in noisy environments.

The effects of L1 orthography on processing an artificial logographic script

To date, studies have focused on the acquisition of alphabetic second languages (L2s) in alphabetic first language (L1) users, demonstrating significant transfer effects. The present study examined the process from a reverse perspective, comparing logographic (Mandarin-Chinese) and alphabetic (English) L1 users in the acquisition of an artificial logographic script, in order to determine whether similar language-specific advantageous transfer effects occurred. English monolinguals, English-French bilinguals and Chinese-English bilinguals learned a small set of symbols in an artificial logographic script and were subsequently tested on their ability to process this script in regard to three main perspectives: L2 reading, L2 working memory (WM), and inner processing strategies. In terms of L2 reading, a lexical decision task on the artificial symbols revealed markedly faster response times in the Chinese-English bilinguals, indicating a logographic transfer effect suggestive of a visual processing advantage. A syntactic decision task evaluated the degree to which the new language was mastered beyond the single word level. No L1-specific transfer effects were found for artificial language strings. In order to investigate visual processing of the artificial logographs further, a series of WM experiments were conducted. Artificial logographs were recalled under concurrent auditory and visuo-spatial suppression conditions to disrupt phonological and visual processing, respectively. No L1-specific transfer effects were found, indicating no visual processing advantage of the Chinese-English bilinguals. However, a bilingual processing advantage was found indicative of a superior ability to control executive functions. In terms of L1 WM, the Chinese-English bilinguals outperformed the alphabetic L1 users when processing L1 words, indicating a language experience-specific advantage. Questionnaire data on the cognitive strategies that were deployed during the acquisition and processing of the artificial logographic script revealed that the Chinese-English bilinguals rated their inner speech as lower than the alphabetic L1 users, suggesting that they were transferring their phonological processing skill set to the acquisition and use of an artificial script. Overall, evidence was found to indicate that language learners transfer specific L1 orthographic processing skills to L2 logographic processing. Additionally, evidence was also found indicating that a bilingual history enhances cognitive performance in L2.

Coding speech signals at very low rates (below 1 kb/s) with high intelligibility

This thesis presents an original approach to parametric speech coding at rates below 1 kbitsjsec, primarily for speech storage applications. Essential processes considered in this research encompass efficient characterization of evolutionary configuration of vocal tract to follow phonemic features with high fidelity, representation of speech excitation using minimal parameters with minor degradation in naturalness of synthesized speech, and finally, quantization of resulting parameters at the nominated rates. For encoding speech spectral features, a new method relying on Temporal Decomposition (TD) is developed which efficiently compresses spectral information through interpolation between most steady points over time trajectories of spectral parameters using a new basis function. The compression ratio provided by the method is independent of the updating rate of the feature vectors, hence allows high resolution in tracking significant temporal variations of speech formants with no effect on the spectral data rate. Accordingly, regardless of the quantization technique employed, the method yields a high compression ratio without sacrificing speech intelligibility. Several new techniques for improving performance of the interpolation of spectral parameters through phonetically-based analysis are proposed and implemented in this research, comprising event approximated TD, near-optimal shaping event approximating functions, efficient speech parametrization for TD on the basis of an extensive investigation originally reported in this thesis, and a hierarchical error minimization algorithm for decomposition of feature parameters which significantly reduces the complexity of the interpolation process. Speech excitation in this work is characterized based on a novel Multi-Band Excitation paradigm which accurately determines the harmonic structure in the LPC (linear predictive coding) residual spectra, within individual bands, using the concept 11 of Instantaneous Frequency (IF) estimation in frequency domain. The model yields aneffective two-band approximation to excitation and computes pitch and voicing with high accuracy as well. New methods for interpolative coding of pitch and gain contours are also developed in this thesis. For pitch, relying on the correlation between phonetic evolution and pitch variations during voiced speech segments, TD is employed to interpolate the pitch contour between critical points introduced by event centroids. This compresses pitch contour in the ratio of about 1/10 with negligible error. To approximate gain contour, a set of uniformly-distributed Gaussian event-like functions is used which reduces the amount of gain information to about 1/6 with acceptable accuracy. The thesis also addresses a new quantization method applied to spectral features on the basis of statistical properties and spectral sensitivity of spectral parameters extracted from TD-based analysis. The experimental results show that good quality speech, comparable to that of conventional coders at rates over 2 kbits/sec, can be achieved at rates 650-990 bits/sec.

Acoustic keyword spotting in speech with applications to data mining

Keyword Spotting is the task of detecting keywords of interest within continu- ous speech. The applications of this technology range from call centre dialogue systems to covert speech surveillance devices. Keyword spotting is particularly well suited to data mining tasks such as real-time keyword monitoring and unre- stricted vocabulary audio document indexing. However, to date, many keyword spotting approaches have su®ered from poor detection rates, high false alarm rates, or slow execution times, thus reducing their commercial viability. This work investigates the application of keyword spotting to data mining tasks. The thesis makes a number of major contributions to the ¯eld of keyword spotting. The ¯rst major contribution is the development of a novel keyword veri¯cation method named Cohort Word Veri¯cation. This method combines high level lin- guistic information with cohort-based veri¯cation techniques to obtain dramatic improvements in veri¯cation performance, in particular for the problematic short duration target word class. The second major contribution is the development of a novel audio document indexing technique named Dynamic Match Lattice Spotting. This technique aug- ments lattice-based audio indexing principles with dynamic sequence matching techniques to provide robustness to erroneous lattice realisations. The resulting algorithm obtains signi¯cant improvement in detection rate over lattice-based audio document indexing while still maintaining extremely fast search speeds. The third major contribution is the study of multiple veri¯er fusion for the task of keyword veri¯cation. The reported experiments demonstrate that substantial improvements in veri¯cation performance can be obtained through the fusion of multiple keyword veri¯ers. The research focuses on combinations of speech background model based veri¯ers and cohort word veri¯ers. The ¯nal major contribution is a comprehensive study of the e®ects of limited training data for keyword spotting. This study is performed with consideration as to how these e®ects impact the immediate development and deployment of speech technologies for non-English languages.

The use of phase in complex spectrum subtraction for robust speech recognition

In this paper we propose a new method for utilising phase information by complementing it with traditional magnitude-only spectral subtraction speech enhancement through Complex Spectrum Subtraction (CSS). The proposed approach has the following advantages over traditional magnitude-only spectral subtraction: (a) it introduces complementary information to the enhancement algorithm; (b) it reduces the total number of algorithmic parameters, and; (c) is designed for improving clean speech magnitude spectra and is therefore suitable for both automatic speech recognition (ASR) and speech perception applications. Oracle-based ASR experiments verify this approach, showing an average of 20% relative word accuracy improvements when accurate estimates of the phase spectrum are available. Based on sinusoidal analysis and assuming stationarity between observations (which is shown to be better approximated as the frame rate is increased), this paper also proposes a novel method for acquiring the phase information called Phase Estimation via Delay Projection (PEDEP). Further oracle ASR experiments validate the potential for the proposed PEDEP technique in ideal conditions. Realistic implementation of CSS with PEDEP shows performance comparable to state of the art spectral subtraction techniques in a range of 15-20 dB signal-to-noise ratio environments. These results clearly demonstrate the potential for using phase spectra in spectral subtractive enhancement applications, and at the same time highlight the need for deriving more accurate phase estimates in a wider range of noise conditions.

Microphone array beamforming approach to blind speech separation

In this paper, we present a microphone array beamforming approach to blind speech separation. Unlike previous beamforming approaches, our system does not require a-priori knowledge of the microphone placement and speaker location, making the system directly comparable other blind source separation methods which require no prior knowledge of recording conditions. Microphone location is automatically estimated using an assumed noise field model, and speaker locations are estimated using cross correlation based methods. The system is evaluated on the data provided for the PASCAL Speech Separation Challenge 2 (SSC2), achieving a word error rate of 58% on the evaluation set.

Low-cost hardware speech enhancement for improved speech recognition in automotive environments

Voice recognition is one of the key enablers to reduce driver distraction as in-vehicle systems become more and more complex. With the integration of voice recognition in vehicles, safety and usability are improved as the driver’s eyes and hands are not required to operate system controls. Whilst speaker independent voice recognition is well developed, performance in high noise environments (e.g. vehicles) is still limited. La Trobe University and Queensland University of Technology have developed a low-cost hardware-based speech enhancement system for automotive environments based on spectral subtraction and delay–sum beamforming techniques. The enhancement algorithms have been optimised using authentic Australian English collected under typical driving conditions. Performance tests conducted using speech data collected under variety of vehicle noise conditions demonstrate a word recognition rate improvement in the order of 10% or more under the noisiest conditions. Currently developed to a proof of concept stage there is potential for even greater performance improvement.

Recognising audio-visual speech in vehicles using the AVICAR database

Interacting with technology within a vehicle environment using a voice interface can greatly reduce the effects of driver distraction. Most current approaches to this problem only utilise the audio signal, making them susceptible to acoustic noise. An obvious approach to circumvent this is to use the visual modality in addition. However, capturing, storing and distributing audio-visual data in a vehicle environment is very costly and difficult. One current dataset available for such research is the AVICAR [1] database. Unfortunately this database is largely unusable due to timing mismatch between the two streams and in addition, no protocol is available. We have overcome this problem by re-synchronising the streams on the phone-number portion of the dataset and established a protocol for further research. This paper presents the first audio-visual results on this dataset for speaker-independent speech recognition. We hope this will serve as a catalyst for future research in this area.

A virtual test assessment tool for second year signal processing students

The development and use of a virtual assessment tool for a signal processing unit is described. It allows students to take a test from anywhere using a web browser to connect to the university server that hosts the test. While student responses are of the multiple choice type, they have to work out problems to arrive at the answer to be entered. CGI programming is used to verify student identification information and record their scores as well as provide immediate feedback after the test is complete. The tool has been used at QUT for the past 3 years and student feedback is discussed. The virtual assessment tool is an efficient alternative to marking written assignment reports that can often take more hours than actual lecture hall contact from a lecturer or tutor. It is especially attractive for very large classes that are now the norm at many universities in the first two years.

Forced landing technologies for unmanned aerial vehicles : towards safer operations

While using unmanned systems in combat is not new, what will be new in the foreseeable future is how such systems are used and integrated in the civilian space. The potential use of Unmanned Aerial Vehicles in civil and commercial applications is becoming a fact, and is receiving considerable attention by industry and the research community. The majority of Unmanned Aerial Vehicles performing civilian tasks are restricted to flying only in segregated space, and not within the National Airspace. The areas that UAVs are restricted to flying in are typically not above populated areas, which in turn are the areas most useful for civilian applications. The reasoning behind the current restrictions is mainly due to the fact that current UAV technologies are not able to demonstrate an Equivalent Level of Safety to manned aircraft, particularly in the case of an engine failure which would require an emergency or forced landing. This chapter will preset and guide the reader through a number of developments that would facilitate the integration of UAVs into the National Airspace. Algorithms for UAV Sense-and-Avoid and Force Landings are recognized as two major enabling technologies that will allow the integration of UAVs in the civilian airspace. The following sections will describe some of the techniques that are currently being tested at the Australian Research Centre for Aerospace Automation (ARCAA), which places emphasis on the detection of candidate landing sites using computer vision, the planning of the descent path trajectory for the UAV, and the decision making process behind the selection of the final landing site.

Improving visual noise insensitivity in small vocabulary audio visual speech recognition applications

Visual noise insensitivity is important to audio visual speech recognition (AVSR). Visual noise can take on a number of forms such as varying frame rate, occlusion, lighting or speaker variabilities. The use of a high dimensional secondary classifier on the word likelihood scores from both the audio and video modalities is investigated for the purposes of adaptive fusion. Preliminary results are presented demonstrating performance above the catastrophic fusion boundary for our confidence measure irrespective of the type of visual noise presented to it. Our experiments were restricted to small vocabulary applications.

The design and development of an undergraduate signal processing laboratory

The School of Electrical and Electronic Systems Engineering of Queensland University of Technology (like many other universities around the world) has recognised the importance of complementing the teaching of signal processing with computer based experiments. A laboratory has been developed to provide a "hands-on" approach to the teaching of signal processing techniques. The motivation for the development of this laboratory was the cliche "What I hear I remember but what I do I understand." The laboratory has been named as the "Signal Computing and Real-time DSP Laboratory" and provides practical training to approximately 150 final year undergraduate students each year. The paper describes the novel features of the laboratory, techniques used in the laboratory based teaching, interesting aspects of the experiments that have been developed and student evaluation of the teaching techniques

Robust speaker verification via fusion of speech and lip modalities

This paper investigates the use of lip information, in conjunction with speech information, for robust speaker verification in the presence of background noise. It has been previously shown in our own work, and in the work of others, that features extracted from a speaker's moving lips hold speaker dependencies which are complementary with speech features. We demonstrate that the fusion of lip and speech information allows for a highly robust speaker verification system which outperforms the performance of either sub-system. We present a new technique for determining the weighting to be applied to each modality so as to optimize the performance of the fused system. Given a correct weighting, lip information is shown to be highly effective for reducing the false acceptance and false rejection error rates in the presence of background noise