Data-driven impostor selection for T-norm score normalisation and the background dataset in SVM-based speaker verification

A data-driven background dataset refinement technique was recently proposed for SVM based speaker verification. This method selects a refined SVM background dataset from a set of candidate impostor examples after individually ranking examples by their relevance. This paper extends this technique to the refinement of the T-norm dataset for SVM-based speaker verification. The independent refinement of the background and T-norm datasets provides a means of investigating the sensitivity of SVM-based speaker verification performance to the selection of each of these datasets. Using refined datasets provided improvements of 13% in min. DCF and 9% in EER over the full set of impostor examples on the 2006 SRE corpus with the majority of these gains due to refinement of the T-norm dataset. Similar trends were observed for the unseen data of the NIST 2008 SRE.

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.

Technology, normalisation and male sex work

Technological change, particularly the growth of the Internet and smart phones, has increased the visibility of male escorts, expanded their client base and diversified the range of venues in which male sex work can take place. Specifically, the Internet has relocated some forms of male sex work away from the street and thereby increased market reach, visibility and access and the scope of sex work advertising. Using the online profiles of 257 male sex workers drawn from six of the largest websites advertising male sexual services in Australia, the role of the Internet in facilitating the normalisation of male sex work is discussed. Specifically we examine how engagement with the sex industry has been reconstituted in term of better informed consumer-seller decisions for both clients and sex workers. Rather than being seen as a ‘deviant’ activity, understood in terms of pathology or criminal activity, male sex work is increasingly presented as an everyday commodity in the market place. In this context, the management of risks associated with sex work has shifted from formalised social control to more informal practices conducted among online communities of clients and sex workers. We discuss the implications for health, legal and welfare responses within an empowerment paradigm.

Document timespan normalisation and understanding temporality for clinical records search

Previous qualitative research has highlighted that temporality plays an important role in relevance for clinical records search. In this study, an investigation is undertaken to determine the effect that the timespan of events within a patient record has on relevance in a retrieval scenario. In addition, based on the standard practise of document length normalisation, a document timespan normalisation model that specifically accounts for timespans is proposed. Initial analysis revealed that in general relevant patient records tended to cover a longer timespan of events than non-relevant patient records. However, an empirical evaluation using the TREC Medical Records track supports the opposite view that shorter documents (in terms of timespan) are better for retrieval. These findings highlight that the role of temporality in relevance is complex and how to effectively deal with temporality within a retrieval scenario remains an open question.

ENEKuS – a key model for managing the transformation of the normalisation of the Basque language in the workplace

The aim of this study is to develop a reference model for intervention in the language processes applied to the transformation of language normalisation within organisations of a socio-economic nature. It is based on the case study of an experience carried out over10 years within a trades’ union confederation, and has pursued a strategy of a basically qualitative research carried out in three stages: 1) undertaking field work through application of action-research methodology, 2) reconstructing experiences following processes of systematisation and conceptualisation of the systematised data, applying methodologies for the Systematisation of Experiences and Grounded Theory, and 3) formulating a model for intervention, applying the Systems Approach methodology. Finally, we identified nine key ideas that make up the conceptual framework for the ENEKuS reference model, which is structured in nine ‘action points', each having an operating sub-model applicable in practice.

Vowel normalisation: Time-domain processing of the internal dynamics of speech

Human listeners can identify vowels regardless of speaker size, although the sound waves for an adult and a child speaking the ’same’ vowel would differ enormously. The differences are mainly due to the differences in vocal tract length (VTL) and glottal pulse rate (GPR) which are both related to body size. Automatic speech recognition machines are notoriously bad at understanding children if they have been trained on the speech of an adult. In this paper, we propose that the auditory system adapts its analysis of speech sounds, dynamically and automatically to the GPR and VTL of the speaker on a syllable-to-syllable basis. We illustrate how this rapid adaptation might be performed with the aid of a computational version of the auditory image model, and we propose that an auditory preprocessor of this form would improve the robustness of speech recognisers.