Efficient real-time face detection for high resolution surveillance applications

This paper presents an efficient face detection method suitable for real-time surveillance applications. Improved efficiency is achieved by constraining the search window of an AdaBoost face detector to pre-selected regions. Firstly, the proposed method takes a sparse grid of sample pixels from the image to reduce whole image scan time. A fusion of foreground segmentation and skin colour segmentation is then used to select candidate face regions. Finally, a classifier-based face detector is applied only to selected regions to verify the presence of a face (the Viola-Jones detector is used in this paper). The proposed system is evaluated using 640 x 480 pixels test images and compared with other relevant methods. Experimental results show that the proposed method reduces the detection time to 42 ms, where the Viola-Jones detector alone requires 565 ms (on a desktop processor). This improvement makes the face detector suitable for real-time applications. Furthermore, the proposed method requires 50% of the computation time of the best competing method, while reducing the false positive rate by 3.2% and maintaining the same hit rate.

Improved foreground detection via block-based classifier cascade with probabilistic decision integration

Background subtraction is a fundamental low-level processing task in numerous computer vision applications. The vast majority of algorithms process images on a pixel-by-pixel basis, where an independent decision is made for each pixel. A general limitation of such processing is that rich contextual information is not taken into account. We propose a block-based method capable of dealing with noise, illumination variations, and dynamic backgrounds, while still obtaining smooth contours of foreground objects. Specifically, image sequences are analyzed on an overlapping block-by-block basis. A low-dimensional texture descriptor obtained from each block is passed through an adaptive classifier cascade, where each stage handles a distinct problem. A probabilistic foreground mask generation approach then exploits block overlaps to integrate interim block-level decisions into final pixel-level foreground segmentation. Unlike many pixel-based methods, ad-hoc postprocessing of foreground masks is not required. Experiments on the difficult Wallflower and I2R datasets show that the proposed approach obtains on average better results (both qualitatively and quantitatively) than several prominent methods. We furthermore propose the use of tracking performance as an unbiased approach for assessing the practical usefulness of foreground segmentation methods, and show that the proposed approach leads to considerable improvements in tracking accuracy on the CAVIAR dataset.

Optimal distribution network reinforcement considering load growth, line loss, and reliability

In this paper, a new comprehensive planning methodology is proposed for implementing distribution network reinforcement. The load growth, voltage profile, distribution line loss, and reliability are considered in this procedure. A time-segmentation technique is employed to reduce the computational load. Options considered range from supporting the load growth using the traditional approach of upgrading the conventional equipment in the distribution network, through to the use of dispatchable distributed generators (DDG). The objective function is composed of the construction cost, loss cost and reliability cost. As constraints, the bus voltages and the feeder currents should be maintained within the standard level. The DDG output power should not be less than a ratio of its rated power because of efficiency. A hybrid optimization method, called modified discrete particle swarm optimization, is employed to solve this nonlinear and discrete optimization problem. A comparison is performed between the optimized solution based on planning of capacitors along with tap-changing transformer and line upgrading and when DDGs are included in the optimization.

Robots move : bootstrapping the development of object representations using sensorimotor coordination

This paper is concerned with the unsupervised learning of object representations by fusing visual and motor information. The problem is posed for a mobile robot that develops its representations as it incrementally gathers data. The scenario is problematic as the robot only has limited information at each time step with which it must generate and update its representations. Object representations are refined as multiple instances of sensory data are presented; however, it is uncertain whether two data instances are synonymous with the same object. This process can easily diverge from stability. The premise of the presented work is that a robot's motor information instigates successful generation of visual representations. An understanding of self-motion enables a prediction to be made before performing an action, resulting in a stronger belief of data association. The system is implemented as a data-driven partially observable semi-Markov decision process. Object representations are formed as the process's hidden states and are coordinated with motor commands through state transitions. Experiments show the prediction process is essential in enabling the unsupervised learning method to converge to a solution - improving precision and recall over using sensory data alone.

Anatomical MR imaging of long bones : comparative performance of MRI at 1.5 T and 3 T

The current gold standard for the design of orthopaedic implants is 3D models of long bones obtained using computed tomography (CT). However, high-resolution CT imaging involves high radiation exposure, which limits its use in healthy human volunteers. Magnetic resonance imaging (MRI) is an attractive alternative for the scanning of healthy human volunteers for research purposes. Current limitations of MRI include difficulties of tissue segmentation within joints and long scanning times. In this work, we explore the possibility of overcoming these limitations through the use of MRI scanners operating at a higher field strength. We quantitatively compare the quality of anatomical MR images of long bones obtained at 1.5 T and 3 T and optimise the scanning protocol of 3 T MRI. FLASH images of the right leg of five human volunteers acquired at 1.5 T and 3 T were compared in terms of signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR). The comparison showed a relatively high CNR and SNR at 3 T for most regions of the femur and tibia, with the exception of the distal diaphyseal region of the femur and the mid diaphyseal region of the tibia. This was accompanied by an ~65% increase in the longitudinal spin relaxation time (T1) of the muscle at 3 T compared to 1.5 T. The results suggest that MRI at 3 T may be able to enhance the segmentability and potentially improve the accuracy of 3D anatomical models of long bones, compared to 1.5 T. We discuss how the total imaging times at 3 T can be kept short while maximising the CNR and SNR of the images obtained.

Cross-Lingual knowledge discovery : Chinese-to-English article linking in Wikipedia

In this paper we examine automated Chinese to English link discovery in Wikipedia and the effects of Chinese segmentation and Chinese to English translation on the hyperlink recommendation. Our experimental results show that the implemented link discovery framework can effectively recommend Chinese-to-English cross-lingual links. The techniques described here can assist bi-lingual users where a particular topic is not covered in Chinese, is not equally covered in both languages, or is biased in one language; as well as for language learning.

SAIVT-QUT@TRECVid 2012 : interactive surveillance event detection

In this paper, we propose an approach which attempts to solve the problem of surveillance event detection, assuming that we know the definition of the events. To facilitate the discussion, we first define two concepts. The event of interest refers to the event that the user requests the system to detect; and the background activities are any other events in the video corpus. This is an unsolved problem due to many factors as listed below: 1) Occlusions and clustering: The surveillance scenes which are of significant interest at locations such as airports, railway stations, shopping centers are often crowded, where occlusions and clustering of people are frequently encountered. This significantly affects the feature extraction step, and for instance, trajectories generated by object tracking algorithms are usually not robust under such a situation. 2) The requirement for real time detection: The system should process the video fast enough in both of the feature extraction and the detection step to facilitate real time operation. 3) Massive size of the training data set: Suppose there is an event that lasts for 1 minute in a video with a frame rate of 25fps, the number of frames for this events is 60X25 = 1500. If we want to have a training data set with many positive instances of the event, the video is likely to be very large in size (i.e. hundreds of thousands of frames or more). How to handle such a large data set is a problem frequently encountered in this application. 4) Difficulty in separating the event of interest from background activities: The events of interest often co-exist with a set of background activities. Temporal groundtruth typically very ambiguous, as it does not distinguish the event of interest from a wide range of co-existing background activities. However, it is not practical to annotate the locations of the events in large amounts of video data. This problem becomes more serious in the detection of multi-agent interactions, since the location of these events can often not be constrained to within a bounding box. 5) Challenges in determining the temporal boundaries of the events: An event can occur at any arbitrary time with an arbitrary duration. The temporal segmentation of events is difficult and ambiguous, and also affected by other factors such as occlusions.

Link discovery for Chinese/English cross-language web information retrieval

Nowadays people heavily rely on the Internet for information and knowledge. Wikipedia is an online multilingual encyclopaedia that contains a very large number of detailed articles covering most written languages. It is often considered to be a treasury of human knowledge. It includes extensive hypertext links between documents of the same language for easy navigation. However, the pages in different languages are rarely cross-linked except for direct equivalent pages on the same subject in different languages. This could pose serious difficulties to users seeking information or knowledge from different lingual sources, or where there is no equivalent page in one language or another. In this thesis, a new information retrieval task—cross-lingual link discovery (CLLD) is proposed to tackle the problem of the lack of cross-lingual anchored links in a knowledge base such as Wikipedia. In contrast to traditional information retrieval tasks, cross language link discovery algorithms actively recommend a set of meaningful anchors in a source document and establish links to documents in an alternative language. In other words, cross-lingual link discovery is a way of automatically finding hypertext links between documents in different languages, which is particularly helpful for knowledge discovery in different language domains. This study is specifically focused on Chinese / English link discovery (C/ELD). Chinese / English link discovery is a special case of cross-lingual link discovery task. It involves tasks including natural language processing (NLP), cross-lingual information retrieval (CLIR) and cross-lingual link discovery. To justify the effectiveness of CLLD, a standard evaluation framework is also proposed. The evaluation framework includes topics, document collections, a gold standard dataset, evaluation metrics, and toolkits for run pooling, link assessment and system evaluation. With the evaluation framework, performance of CLLD approaches and systems can be quantified. This thesis contributes to the research on natural language processing and cross-lingual information retrieval in CLLD: 1) a new simple, but effective Chinese segmentation method, n-gram mutual information, is presented for determining the boundaries of Chinese text; 2) a voting mechanism of name entity translation is demonstrated for achieving a high precision of English / Chinese machine translation; 3) a link mining approach that mines the existing link structure for anchor probabilities achieves encouraging results in suggesting cross-lingual Chinese / English links in Wikipedia. This approach was examined in the experiments for better, automatic generation of cross-lingual links that were carried out as part of the study. The overall major contribution of this thesis is the provision of a standard evaluation framework for cross-lingual link discovery research. It is important in CLLD evaluation to have this framework which helps in benchmarking the performance of various CLLD systems and in identifying good CLLD realisation approaches. The evaluation methods and the evaluation framework described in this thesis have been utilised to quantify the system performance in the NTCIR-9 Crosslink task which is the first information retrieval track of this kind.

Users segmentations for recommendation

Traditional recommendation methods provide recommendations equally to all users. In this paper, a segmentation method using the Gaussian Mixture Model (GMM) is proposed to customize users’ needs in order to offer a specific recommendation strategy to each segment. Experiment is conducted using a live online dating network data.

Speaker diarization : "who spoke when"

Speaker diarization is the process of annotating an input audio with information that attributes temporal regions of the audio signal to their respective sources, which may include both speech and non-speech events. For speech regions, the diarization system also specifies the locations of speaker boundaries and assign relative speaker labels to each homogeneous segment of speech. In short, speaker diarization systems effectively answer the question of ‘who spoke when’. There are several important applications for speaker diarization technology, such as facilitating speaker indexing systems to allow users to directly access the relevant segments of interest within a given audio, and assisting with other downstream processes such as summarizing and parsing. When combined with automatic speech recognition (ASR) systems, the metadata extracted from a speaker diarization system can provide complementary information for ASR transcripts including the location of speaker turns and relative speaker segment labels, making the transcripts more readable. Speaker diarization output can also be used to localize the instances of specific speakers to pool data for model adaptation, which in turn boosts transcription accuracies. Speaker diarization therefore plays an important role as a preliminary step in automatic transcription of audio data. The aim of this work is to improve the usefulness and practicality of speaker diarization technology, through the reduction of diarization error rates. In particular, this research is focused on the segmentation and clustering stages within a diarization system. Although particular emphasis is placed on the broadcast news audio domain and systems developed throughout this work are also trained and tested on broadcast news data, the techniques proposed in this dissertation are also applicable to other domains including telephone conversations and meetings audio. Three main research themes were pursued: heuristic rules for speaker segmentation, modelling uncertainty in speaker model estimates, and modelling uncertainty in eigenvoice speaker modelling. The use of heuristic approaches for the speaker segmentation task was first investigated, with emphasis placed on minimizing missed boundary detections. A set of heuristic rules was proposed, to govern the detection and heuristic selection of candidate speaker segment boundaries. A second pass, using the same heuristic algorithm with a smaller window, was also proposed with the aim of improving detection of boundaries around short speaker segments. Compared to single threshold based methods, the proposed heuristic approach was shown to provide improved segmentation performance, leading to a reduction in the overall diarization error rate. Methods to model the uncertainty in speaker model estimates were developed, to address the difficulties associated with making segmentation and clustering decisions with limited data in the speaker segments. The Bayes factor, derived specifically for multivariate Gaussian speaker modelling, was introduced to account for the uncertainty of the speaker model estimates. The use of the Bayes factor also enabled the incorporation of prior information regarding the audio to aid segmentation and clustering decisions. The idea of modelling uncertainty in speaker model estimates was also extended to the eigenvoice speaker modelling framework for the speaker clustering task. Building on the application of Bayesian approaches to the speaker diarization problem, the proposed approach takes into account the uncertainty associated with the explicit estimation of the speaker factors. The proposed decision criteria, based on Bayesian theory, was shown to generally outperform their non- Bayesian counterparts.

Introducing Masculinity Studies to Information Systems research : the case of Gaydar

I believe that studies of men's gendered experiences of information systems are needed. In order to support this claim, I introduce the area of Masculinity Studies to Information Systems research and, using this, present an exploratory analysis of an internet dating website for gay men – Gaydar. The information system, which forms part of the Gaydar community, is shown to shape, and be shaped by the members as they accept and challenge aspects of it as related to their identities. In doing this, I show how the intertwined processes of information systems development and use contribute to the creation of diverse interpretations of masculinity within a group of men. In sum, my analysis highlights different kinds of men and different versions of masculinity that can sometimes be associated with different experiences of information systems. The implications of this work centre on the need to expand our knowledge of men's gendered experiences with information systems, to reflect upon processes of technology facilitated categorisation and to consider the influences that contribute to the roll out of particular software features along with the underlying rationales for market segmentation in the software and software-based services industries.

Urban transit quality of service : user perception and behaviour

Despite its potential multiple contributions to sustainable policy objectives, urban transit is generally not widely used by the public in terms of its market share compared to that of automobiles, particularly in affluent societies with low-density urban forms like Australia. Transit service providers need to attract more people to transit by improving transit quality of service. The key to cost-effective transit service improvements lies in accurate evaluation of policy proposals by taking into account their impacts on transit users. If transit providers knew what is more or less important to their customers, they could focus their efforts on optimising customer-oriented service. Policy interventions could also be specified to influence transit users’ travel decisions, with targets of customer satisfaction and broader community welfare. This significance motivates the research into the relationship between urban transit quality of service and its user perception as well as behaviour. This research focused on two dimensions of transit user’s travel behaviour: route choice and access arrival time choice. The study area chosen was a busy urban transit corridor linking Brisbane central business district (CBD) and the St. Lucia campus of The University of Queensland (UQ). This multi-system corridor provided a ‘natural experiment’ for transit users between the CBD and UQ, as they can choose between busway 109 (with grade-separate exclusive right-of-way), ordinary on-street bus 412, and linear fast ferry CityCat on the Brisbane River. The population of interest was set as the attendees to UQ, who travelled from the CBD or from a suburb via the CBD. Two waves of internet-based self-completion questionnaire surveys were conducted to collect data on sampled passengers’ perception of transit service quality and behaviour of using public transit in the study area. The first wave survey is to collect behaviour and attitude data on respondents’ daily transit usage and their direct rating of importance on factors of route-level transit quality of service. A series of statistical analyses is conducted to examine the relationships between transit users’ travel and personal characteristics and their transit usage characteristics. A factor-cluster segmentation procedure is applied to respodents’ importance ratings on service quality variables regarding transit route preference to explore users’ various perspectives to transit quality of service. Based on the perceptions of service quality collected from the second wave survey, a series of quality criteria of the transit routes under study was quantitatively measured, particularly, the travel time reliability in terms of schedule adherence. It was proved that mixed traffic conditions and peak-period effects can affect transit service reliability. Multinomial logit models of transit user’s route choice were estimated using route-level service quality perceptions collected in the second wave survey. Relative importance of service quality factors were derived from choice model’s significant parameter estimates, such as access and egress times, seat availability, and busway system. Interpretations of the parameter estimates were conducted, particularly the equivalent in-vehicle time of access and egress times, and busway in-vehicle time. Market segmentation by trip origin was applied to investigate the difference in magnitude between the parameter estimates of access and egress times. The significant costs of transfer in transit trips were highlighted. These importance ratios were applied back to quality perceptions collected as RP data to compare the satisfaction levels between the service attributes and to generate an action relevance matrix to prioritise attributes for quality improvement. An empirical study on the relationship between average passenger waiting time and transit service characteristics was performed using the service quality perceived. Passenger arrivals for services with long headways (over 15 minutes) were found to be obviously coordinated with scheduled departure times of transit vehicles in order to reduce waiting time. This drove further investigations and modelling innovations in passenger’ access arrival time choice and its relationships with transit service characteristics and average passenger waiting time. Specifically, original contributions were made in formulation of expected waiting time, analysis of the risk-aversion attitude to missing desired service run in the passengers’ access time arrivals’ choice, and extensions of the utility function specification for modelling passenger access arrival distribution, by using complicated expected utility forms and non-linear probability weighting to explicitly accommodate the risk of missing an intended service and passenger’s risk-aversion attitude. Discussions on this research’s contributions to knowledge, its limitations, and recommendations for future research are provided at the concluding section of this thesis.

A gender comparison of food shopper typologies

This research segments male and female food shoppers based on store and product attribute evaluations, enjoyment, planning, and shopping responsibility. A rich profile for each segment is developed from a sample of 580 primary food shoppers. Gender comparisons are operationalized and these developed food shopper typologies are contrasted against earlier works. Factor analysis, cluster analysis, and ANOVA were employed to develop specific segments of shoppers. This is the first study to identify specific groups of male food shoppers and contrast these to traditional female shoppers and earlier typographical works. The research provides the basis for further cross-cultural, cross-contextual comparative studies.

Utilizing least significant bit-planes of RONI pixels for medical image watermarking

We propose a computationally efficient image border pixel based watermark embedding scheme for medical images. We considered the border pixels of a medical image as RONI (region of non-interest), since those pixels have no or little interest to doctors and medical professionals irrespective of the image modalities. Although RONI is used for embedding, our proposed scheme still keeps distortion at a minimum level in the embedding region using the optimum number of least significant bit-planes for the border pixels. All these not only ensure that a watermarked image is safe for diagnosis, but also help minimize the legal and ethical concerns of altering all pixels of medical images in any manner (e.g, reversible or irreversible). The proposed scheme avoids the need for RONI segmentation, which incurs capacity and computational overheads. The performance of the proposed scheme has been compared with a relevant scheme in terms of embedding capacity, image perceptual quality (measured by SSIM and PSNR), and computational efficiency. Our experimental results show that the proposed scheme is computationally efficient, offers an image-content-independent embedding capacity, and maintains a good image quality

Making the train : an examination of five perspectives of rail access and how they resonate with existing and potential passengers

Rail operators recognize a need to increase ridership in order to improve the economic viability of rail service, and to magnify the role that rail travel plays in making cities feel liveable. This study extends previous research that used cluster analysis with a small sample of rail passengers to identify five salient perspectives of rail access (Zuniga et al, 2013). In this project stage, we used correlation techniques to determine how those perspectives would resonate with two larger study populations, including a relatively homogeneous sample of university students in Brisbane, Australia and a diverse sample of rail passengers in Melbourne, Australia. Findings from Zuniga et al. (2013) described a complex typology of current passengers that was based on respondents’ subjective attitudes and perceptions rather than socio-demographic or travel behaviour characteristics commonly used for segmentation analysis. The typology included five qualitative perspectives of rail travel. Based on the transport accessibility literature, we expected to find that perspectives from that study emphasizing physical access to rail stations would be shared by current and potential rail passengers who live further from rail stations. Other perspectives might be shared among respondents who live nearby, since the relevance of distance would be diminished. The population living nearby would thus represent an important target group for increasing ridership, since making rail travel accessible to them does not require expansion of costly infrastructure such as new lines or stations. By measuring the prevalence of each perspective in a larger respondent pool, results from this study provide insight into the typical socio-demographic and travel behaviour characteristics that correspond to each perspective of intra-urban rail travel. In several instances, our quantitative findings reinforced Zuniga et al.’s (2013) qualitative descriptions of passenger types, further validating the original research. This work may directly inform rail operators’ approach to increasing ridership through marketing and improvements to service quality and station experience. Operators in other parts of Australia and internationally may also choose to replicate the study locally, to fine-tune understanding of diverse customer bases. Developing regional and international collaboration would provide additional opportunities to evaluate and benchmark service and station amenities as they address the various access dimensions.