Robust speech recognition using speech enhancement

Automatic Speech Recognition (ASR) has matured into a technology which is becoming more common in our everyday lives, and is emerging as a necessity to minimise driver distraction when operating in-car systems such as navigation and infotainment. In “noise-free” environments, word recognition performance of these systems has been shown to approach 100%, however this performance degrades rapidly as the level of background noise is increased. Speech enhancement is a popular method for making ASR systems more ro- bust. Single-channel spectral subtraction was originally designed to improve hu- man speech intelligibility and many attempts have been made to optimise this algorithm in terms of signal-based metrics such as maximised Signal-to-Noise Ratio (SNR) or minimised speech distortion. Such metrics are used to assess en- hancement performance for intelligibility not speech recognition, therefore mak- ing them sub-optimal ASR applications. This research investigates two methods for closely coupling subtractive-type enhancement algorithms with ASR: (a) a computationally-efficient Mel-filterbank noise subtraction technique based on likelihood-maximisation (LIMA), and (b) in- troducing phase spectrum information to enable spectral subtraction in the com- plex frequency domain. Likelihood-maximisation uses gradient-descent to optimise parameters of the enhancement algorithm to best fit the acoustic speech model given a word se- quence known a priori. Whilst this technique is shown to improve the ASR word accuracy performance, it is also identified to be particularly sensitive to non-noise mismatches between the training and testing data. Phase information has long been ignored in spectral subtraction as it is deemed to have little effect on human intelligibility. In this work it is shown that phase information is important in obtaining highly accurate estimates of clean speech magnitudes which are typically used in ASR feature extraction. Phase Estimation via Delay Projection is proposed based on the stationarity of sinusoidal signals, and demonstrates the potential to produce improvements in ASR word accuracy in a wide range of SNR. Throughout the dissertation, consideration is given to practical implemen- tation in vehicular environments which resulted in two novel contributions – a LIMA framework which takes advantage of the grounding procedure common to speech dialogue systems, and a resource-saving formulation of frequency-domain spectral subtraction for realisation in field-programmable gate array hardware. The techniques proposed in this dissertation were evaluated using the Aus- tralian English In-Car Speech Corpus which was collected as part of this work. This database is the first of its kind within Australia and captures real in-car speech of 50 native Australian speakers in seven driving conditions common to Australian environments.

H-house

The author's approach to the problems associated with building in bushfire prone landscapes comes from 12 years of study of the biophysical and cultural landscapes in the Great Southern Region of Western Australia - research which resulted in the design and construction of the H-house at Bremer Bay. The house was developed using a 'ground up' approach whereby Dr Weir conducted topographical surveys and worked with a local botanist and a bushfire risk consultant to ascertain the level of threat that fire presented to this particular site. The intention from the outset however, was not to design a bushfire resistant house per se, but to develop a design which would place the owners in close proximity to the highly biodiverse heath vegetation of their site. The research aim was to find ways - through architectural design-to link the patterns of usage of the house with other site specific conditions related to the prevailing winds, solar orientation and seasonal change. The H-house has a number of features which increase the level of bushfire safety. These include: Fire rated roller shutters (tested by the CSIRO for ember attack and radiant heat), Fire resistant double glazing (on windows not protected by the shutters), Fibre-cement sheet cladding of the underside of the elevated timber floor structure, Manually operated high pressure sprinkler system on exposed timber decks, A fire refuge (an enlarged laundry, shower area) within the house with a dedicated cabinet for fire fighting equipment) and A low pressure solar powered domestic water supply system.

Bushfire responsive architecture : strategies for reconciling bushfire and biodiversity with daily life

My perspective on the problems associated with building in bushfire prone landscapes comes from 12 years of study of the biophysical and cultural landscapes in the Great Southern Region of WA which resulted in the design and construction of the ‘Hhouse’ at Bremer Bay. The house was developed using a ‘ground up’ approach whereby I conducted a topographical survey and worked with a local botanist and a bushfire risk consultant to ascertain the level of threat that fire presented to this particular site. My intention from the outset however, was not to design a bushfire resistant house per se, but to develop a design which would place the owners in close proximity to the highly biodiverse heath vegetation of the site. I was also seeking a means—through architectural design—of linking the patterns of usage of the house with other site specific conditions related to the prevailing winds, solar orientation and seasonal change.

Performance of reinforced concrete columns under the vehicular impacts

Frontal columns in buildings and columns in car parks are vulnerable to vehicular impacts. This paper treats the impact response of such concrete columns under vehicular loads and the use of polymer wrap to enhance their impact capacity. Comprehensive dynamic computer simulation techniques are used along with strain rate effects and hour glass control to evaluate the impact response. Results indicate the effectiveness of wraps in enhancing the impact capacity of these columns.

The social role of design on collaborative destination branding : creating a new journey, a new story for the Waterfall Way, New South Wales, Australia.

This paper suggests that collaborative design can be an effective tool to promote social change. A co-design methodology and the results of its application in branding the Waterfall Way (New South Wales, Australia) as an eco- and nature-based tourism destination are presented as an example. The co-design exercise actively involved stakeholders in all stages of the design process, harnessing local tacit knowledge in relation to communication design, stimulating reflection upon what is special about the places, and consequently reinforcing a sense of belonging and the environmental and cultural conservation of place. The achieved results reflect the involvement and ownership of the community towards the design process. However, the application of a collaborative brand design methodology produced more than just a destination brand that is attractive to visitors, in line with local values, ways of living and the environment. It helped to catalyse a social network around tourism, triggering self-organising activity amongst stakeholders, who started to liaise with each other around the emergent regional identity - represented by the new brand they created together. The Waterfall Way branding process is a good example of social construction of shared understanding in and through design, showing that design exercises can have a significant social impact not only on the final product, but also on the realities of people involved in the process.

Steering towards pinnacle of success : the NEC tenets in Australian construction contracting

Contractual relationships have become increasingly strained in recent years in the construction industry result in the use of the judicial system for the settlement of contractual disagreements. Why is this so? Evidence from anecdotes suggest that the lack of capacity amongst owners and contractors to carry out a contract using a good practice approach during the construction of a project contribute to the occurrence of conflicts, losses, deficient contractual relationships and poor performance of the construction work. Recognizing that current forms of contract in use today perpetuate a legacy of construction problems, we are conducting explanatory research to examine whether the widely publicized benefits of New Engineering Contract (NEC) could be realized in the Australian construction industry. This paper outlines a research agenda that will help shed light on how contract forms are able to be used as a mechanism to ensure construction projects are delivered successfully whilst also meeting the goals of multiple stakeholders. Understanding the Critical Success Factors (CSFs), commonly used construction contracts and the NEC system can help us address some of these issues. However, there are gaps in the validation of the benefits of NEC and its link with project success. We identify some of these gaps and propose a methodology by which to gain insights into this phenomenon. Keywords: Project Success, Construction Contracting, New Engineering Contract (NEC)

Responding to sustainability challenge and cost implications in highway construction projects

Highway construction often requires a significant capital input; therefore it often causes serious financial implications for developers, owners and operators. The recent industry-wide focus on sustainability has added a new dimension to the evaluation of highway projects, particularly on the economical scale of ‘going green’. Comprehensive analysis of the whole-of-life highway development that responds to sustainability challenges is one of the primary concerns for stakeholders. Principles of engineering economics and life cycle costing have been used to determine the incremental capacity investments for highway projects. However, the consideration of costs and issues associated with sustainability is still very limited in current studies on highway projects. Previous studies have identified that highway project investments are primarily concerned with direct market costs that can be quantified through life cycle costing analysis (LCCA). But they tend to ignore costs that are difficult to calculate, as those related to environmental and social elements. On a more positive note, these studies proved that the inclusion of such costs is an essential part of the overall development investment and a primary concern for decision making by the stakeholders. This paper discusses a research attempt to identify and categorise sustainability cost elements for highway projects. Through questionnaire survey, a set of sustainability cost elements on highway projects has been proposed. These cost elements are incorporated into the extension of some of the existing Life Cycle Costing Analysis (LCCA) models in order to produce a holistic financial picture of the highway project. It is expected that a new LCCA model will be established to serve as a suitable tool for decision making for highway project stakeholders.

Measuring performance of municipal real estate : softening the impact of financial crisis to local governments

Asset management in local government is an emerging discipline and over a decade has become a crucial aspect towards a more efficient and effective organisation. One crucial feature in the public asset management is performance measurement toward the public real estates. This measurement critically at the important component of public wealth and seeks to apply a standard of economic efficiency and effective organisational management especially in such global financial crisis condition. This paper aims to identify global economic crisis effect and proposes alternative solution for local governments to softening the impact of the crisis to the local governments organisation. This study found that the most suitable solution for local government to solve the global economic crisis in Indonesia is application of performance measurement in its asset management. Thus, it is important to develop performance measurement system in local government asset management process. This study provides suggestions from published documents and literatures. The paper also discusses the elements of public real estate performance measurement. The measurement of performance has become an essential component of the strategic thinking of assets owners and managers. Without having a formal measurement system for performance, it is difficult to plan, control and improve local government real estate management system. A close look at best practices in public sectors reveals that in most cases these practices were transferred from private sector reals estate management under the direction of real estate experts retained by government. One of the most significant advances in government property performance measurement resulted from recognition that the methodology used by private sector, non real estate corporations for managing their real property offered a valuable prototype for local governments. In general, there are two approaches most frequently used to measure performance of public organisations. Those are subjective and objective measures. Finally, findings from this study provides useful input for the local government policy makers, scholars and asset management practitioners to establish a public real estate performance measurement system toward more efficient and effective local governments in managing their assets as well as increasing public services quality in order to soften the impact of global financial crisis.

Mapping a suburb with a single camera using a biologically inspired SLAM system

This paper describes a biologically inspired approach to vision-only simultaneous localization and mapping (SLAM) on ground-based platforms. The core SLAM system, dubbed RatSLAM, is based on computational models of the rodent hippocampus, and is coupled with a lightweight vision system that provides odometry and appearance information. RatSLAM builds a map in an online manner, driving loop closure and relocalization through sequences of familiar visual scenes. Visual ambiguity is managed by maintaining multiple competing vehicle pose estimates, while cumulative errors in odometry are corrected after loop closure by a map correction algorithm. We demonstrate the mapping performance of the system on a 66 km car journey through a complex suburban road network. Using only a web camera operating at 10 Hz, RatSLAM generates a coherent map of the entire environment at real-time speed, correctly closing more than 51 loops of up to 5 km in length.

Vulnerability of axially loaded columns subjected to transverse impact loads

Increased industrialisation has brought to the forefront the susceptibility of concrete columns in both buildings and bridges to vehicle impacts. Accurate vulnerability assessments are crucial in the design process due to possible catastrophic nature of the failures that can cause. This paper reports on research undertaken to investigate the impact capacity of the columns of low to medium raised building designed according to Australian Standards. Numerical simulation techniques were used in the process and validation was done by using experimental results published in the literature. The investigation thus far has confirmed that vulnerability of typical columns in five story buildings located in urban areas to medium velocity car impacts and hence these columns need to be re-designed (if possible) or retrofitted. In addition, accuracy of the simplified method presented in EN 1991 to quantify the impact damage was scrutinised. A simplified concept to assess the damage due to all collisions modes was introduced. The research information will be extended to generate a common data base to assess the vulnerability of columns in urban areas against new generation of vehicles.

The rediscovery of traditional urbanism at Melrose Arch

Contemporary urban form, particularly in the cities of South Africa, lacks distinction and quality. The majority of developments are conceived as private and dislocated initiatives, surveiled enclaves with gated access being the only conduit to the outside world. Any concern for a positive contribution to the matrix of public activity is seldom a consideration. The urban form responds to the perception that traffic systems are paramount to the successful flux of the city in satisfying the escalating demands of vehicular movement. In contrast many of the urban centres around the world, the great historical centres of Europe, Americas and the Sub-Continent are admired and considered the ultimate models in urban experience. The colonnades, bazaars and boulevards hosting an abundance of street activity are the characteristics of such centres and are symptomatic of a city growth based on pedestrian movement patterns, an urbanism supportative of human interaction and exchange, a form which has nurtured the existence of a public realm. Through the understanding of the principles of traditional urbanism we may learn that the modernist paradigm of a contemporary suburbia has resulted in disconnected and separate land uses with isolated districts where a reliance on the car is essential rather than optional.

A rights management approach to securing data distribution in coalitions

As network capacity has increased over the past decade, individuals and organisations have found it increasingly appealing to make use of remote services in the form of service-oriented architectures and cloud computing services. Data processed by remote services, however, is no longer under the direct control of the individual or organisation that provided the data, leaving data owners at risk of data theft or misuse. This paper describes a model by which data owners can control the distribution and use of their data throughout a dynamic coalition of service providers using digital rights management technology. Our model allows a data owner to establish the trustworthiness of every member of a coalition employed to process data, and to communicate a machine-enforceable usage policy to every such member.

An application of Keller’s Brand Equity Model in a B2B context

Powerful brands create meaningful images in the minds of customers (Keller, 1993). A strong brand image and reputation enhances differentiation and has a positive influence on buying behaviour (Gordon et al., 1993; McEnally and de Chernatony, 1999). While the power of branding is widely acknowledged in consumer markets, the nature and importance of branding in industrial markets remains under-researched. Many business-to-business (B2B) strategists have claimed brand-building belongs in the consumer realm. They argue that industrial products do not need branding as it is confusing and adds little value to functional products (Collins, 1977; Lorge, 1998; Saunders and Watt, 1979). Others argue that branding and the concept of brand equity however are increasingly important in industrial markets, because it has been shown that what a brand means to a buyer can be a determining factor in deciding between industrial purchase alternatives (Aaker, 1991). In this context, it is critical for suppliers to initiate and sustain relationships due to the small number of potential customers (Ambler, 1995; Webster and Keller, 2004). To date however, there is no model available to assist B2B marketers in identifying and measuring brand equity. In this paper, we take a step in that direction by operationalising and empirically testing a prominent brand equity model in a B2B context. This makes not only a theoretical contribution by advancing branding research, but also addresses a managerial need for information that will assist in the assessment of industrial branding efforts.

Audio visual automatic speech recognition in vehicles

Acoustically, car cabins are extremely noisy and as a consequence, existing audio-only speech recognition systems, for voice-based control of vehicle functions such as the GPS based navigator, perform poorly. Audio-only speech recognition systems fail to make use of the visual modality of speech (eg: lip movements). As the visual modality is immune to acoustic noise, utilising this visual information in conjunction with an audio only speech recognition system has the potential to improve the accuracy of the system. The field of recognising speech using both auditory and visual inputs is known as Audio Visual Speech Recognition (AVSR). Continuous research in AVASR field has been ongoing for the past twenty-five years with notable progress being made. However, the practical deployment of AVASR systems for use in a variety of real-world applications has not yet emerged. The main reason is due to most research to date neglecting to address variabilities in the visual domain such as illumination and viewpoint in the design of the visual front-end of the AVSR system. In this paper we present an AVASR system in a real-world car environment using the AVICAR database [1], which is publicly available in-car database and we show that the use of visual speech conjunction with the audio modality is a better approach to improve the robustness and effectiveness of voice-only recognition systems in car cabin environments.

Sustainable infrastructure assets : capability for infrastructure capacity management

Building for a sustainable environment requires sustainable infrastructure assets. Infrastructure capacity management is the process of ensuring optimal provision of such infrastructure assets. Effectiveness in this process will enable the infrastructure asset owners and its stakeholders to receive full value on their investment. Business research has shown that an organisation can only achieve business value when it has the right capabilities. This paradigm can also be applied to infrastructure capacity management. With limited access to resources, the challenge for infrastructure organisations is to identify and develop core capabilities to enable infrastructure capacity management. This chapter explores the concept of capability and identifies the core capability needed in infrastructure capacity management. Through a case study of the Port of Brisbane, this chapter shows that infrastructure organisations must develop their intelligence gathering capability to effectively manage the capacity of their infrastructure assets.