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.

Automatic spoken language identification utilizing acoustic and phonetic speech information

Automatic spoken Language Identi¯cation (LID) is the process of identifying the language spoken within an utterance. The challenge that this task presents is that no prior information is available indicating the content of the utterance or the identity of the speaker. The trend of globalization and the pervasive popularity of the Internet will amplify the need for the capabilities spoken language identi¯ca- tion systems provide. A prominent application arises in call centers dealing with speakers speaking di®erent languages. Another important application is to index or search huge speech data archives and corpora that contain multiple languages. The aim of this research is to develop techniques targeted at producing a fast and more accurate automatic spoken LID system compared to the previous National Institute of Standards and Technology (NIST) Language Recognition Evaluation. Acoustic and phonetic speech information are targeted as the most suitable fea- tures for representing the characteristics of a language. To model the acoustic speech features a Gaussian Mixture Model based approach is employed. Pho- netic speech information is extracted using existing speech recognition technol- ogy. Various techniques to improve LID accuracy are also studied. One approach examined is the employment of Vocal Tract Length Normalization to reduce the speech variation caused by di®erent speakers. A linear data fusion technique is adopted to combine the various aspects of information extracted from speech. As a result of this research, a LID system was implemented and presented for evaluation in the 2003 Language Recognition Evaluation conducted by the NIST.

The effectiveness of a short-term group music therapy intervention for parents who have a child with a disability

The relationship between the quality of parent-child interactions and positive child developmental trajectories is well established (Guralnick, 2006; Shonkoff & Meissels, 2000; Zubrick et al., 2008). However, a range of parental, family, and socio-economic factors can pose risks to parents’ capacity to participate in quality interactions with their children. In particular, families with a child with a disability have been found to have higher levels of parenting stress, and are more likely to experience economic disadvantage, as well as social isolation. The importance of early interventions to promote positive parenting and child development for these families is widely recognised (Shonkoff & Meissels, 2000). However, to date, there is a lack of evidence about the effectiveness of early parenting programs for families who have a young child with a disability. This thesis investigates the impact of a music therapy parenting program, Sing & Grow, on 201 parent-child dyads who attended programs specifically targeted to parents who had a young child with a disability. Sing & Grow is an Australian national early parenting intervention funded by the Australian Government Department of Families, Housing, Community Services and Indigenous Affairs and delivered by Playgroup Queensland. It is designed and delivered by Registered Music Therapists for families with children aged from birth to three years. It aims to improve parenting skills and confidence, improve family functioning (positive parent-child interactions), enhance child development, and provide social networking opportunities to socially isolated families. The intervention targets a range of families in circumstances that have the potential to impact negatively on family functioning. This thesis uses data from the National Evaluation Study of Sing & Grow from programs which were targeted at families who had a young child with a disability. Three studies were conducted to address the objectives of this thesis. Study 1 examines the effects of the Sing & Grow intervention on parent reported pre and post parent mental health, parenting confidence, parenting skills, and child development, and other parent reported outcomes including social support, use of intervention resources, satisfaction with the intervention and perceived benefits of and barriers to participation. Significant improvements from pre to post were found for parent mental health and parent reported child communication and social skills, along with evidence that parents were very satisfied with the program and that it brought social benefits to families. Study 2 explored the pre to post effects of the intervention on children’s developmental skills and parent-child interactions using observational ratings made by clinicians. Significant pre to post improvements were found for parenting sensitivity, parental engagement with child and acceptance of child as well as for child responsiveness to parent, interest, and participation in the intervention, and social skills. Study 3 examined the nature of child and family characteristics that predicted better outcomes for families while taking account of the level of participation in the program. An overall outcome index was calculated and served as the dependent variable in a logistic regression analysis. Families who attended six or more sessions and mothers who had not completed high school were more likely to have higher outcome scores at post intervention than those who attended fewer sessions and those with more educated mothers respectively. The findings of this research indicate that the intervention had a positive impact on participants’ mental health, parenting behaviours and child development and that level of attendance was associated with better outcomes. There was also evidence that the program reached its target of high risk families (i.e., families in which mothers had lower educational levels) and that for these families better outcomes were achieved. There were also indications that the program was accessible and highly regarded by families and that it promoted social connections for participants. A theoretical model of how the intervention is currently working for families is proposed to explain the connections between early parenting, child development and maternal wellbeing. However, more research is required to further elucidate the mechanisms by which the intervention creates change for families. This research presents promising evidence that a short term group music therapy program can elicit important therapeutic benefits for families who have a child with a disability.

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.