Extraction and classification of self-consumable sport video highlights using generic HMM

This paper aims to automatically extract and classify self-consumable sport video highlights. For this purpose, we will emphasize the benefits of using play-break sequences as the effective inputs for HMMbased classifier. HMM is used to model the stochastic pattern of high-level states during specific sport highlights which correspond to the sequence of generic audio-visual measurements extracted from raw video data. This paper uses soccer as the domain study, focusing on the extraction and classification of goal, shot and foul highlights. The experiment work which uses183 play-break sequences from 6 soccer matches will be presented to demonstrate the performance of our proposed scheme.

Towards universal and statistical-driven heuristics for automatic classification of sports video events

Researchers worldwide have been actively seeking for the most robust and powerful solutions to detect and classify key events (or highlights) in various sports domains. Most approaches have employed manual heuristics that model the typical pattern of audio-visual features within particular sport events To avoid manual observation and knowledge, machine-learning can be used as an alternative approach. To bridge the gaps between these two alternatives, an attempt is made to integrate statistics into heuristic models during highlight detection in our investigation. The models can be designed with a modest amount of domain-knowledge, making them less subjective and more robust for different sports. We have also successfully used a universal scope of detection and a standard set of features that can be applied for different sports that include soccer, basketball and Australian football. An experiment on a large dataset of sport videos, with a total of around 15 hours, has demonstrated the effectiveness and robustness of our
aIlgorithms.

Using decision-tree to automatically construct learned-heuristics for events classification in sports video

Automatic events classification is an essential requirement for constructing an effective sports video summary. It has become a well-known theory that the high-level semantics in sport video can be “computationally interpreted” based on the occurrences of specific audio and visual features which can be extracted automatically. State-of-the-art solutions for features-based event classification have only relied on either manual-knowledge based heuristics or machine learning. To bridge the gaps, we have successfully combined the two approaches by using learning-based heuristics. The heuristics are constructed automatically using decision tree while manual supervision is only required to check the features and highlight contained in each training segment. Thus, fully automated construction of classification system for sports video events has been achieved. A comprehensive experiment on 10 hours video dataset, with five full-match soccer and five full-match basketball videos, has demonstrated the effectiveness/robustness of our algorithms.

Two stage partial classification for inconsistent and imbalanced classes

When deriving classification rules for a non-symmetric database with a binary target class, it is common practice to generate rules for the majority class, then any object which is not covered by a rule of suitable accuracy is by default given the minority class prediction. However, in the case where misclassification costs for the minority class significantly outweigh those of the majority class, this may mean that there are still costly incorrect predictions. We examine the capability of an evolutionary algorithm to detect these potentially costly misclassifications.

Bayesian classification of catchments using spatial data: a first step to improved modelling of catchment effects on stream ecological condition

A major challenge facing freshwater ecologists and managers is the development of models that link stream ecological condition to catchment scale effects, such as land use. Previous attempts to make such models have followed two general approaches. The bottom-up approach employs mechanistic models, which can quickly become too complex to be useful. The top-down approach employs empirical models derived from large data sets, and has often suffered from large amounts of unexplained variation in stream condition.

We believe that the lack of success of both modelling approaches may be at least partly explained by scientists considering too wide a breadth of catchment type. Thus, we believe that by stratifying large sets of catchments into groups of similar types prior to modelling, both types of models may be improved. This paper describes preliminary work using a Bayesian classification software package, ‘Autoclass’ (Cheeseman and Stutz 1996) to create classes of catchments within the Murray Darling Basin based on physiographic data.

Autoclass uses a model-based classification method that employs finite mixture modelling and trades off model fit versus complexity, leading to a parsimonious solution. The software provides information on the posterior probability that the classification is ‘correct’ and also probabilities for alternative classifications. The importance of each attribute in defining the individual classes is calculated and presented, assisting description of the classes. Each case is ‘assigned’ to a class based on membership probability, but the probability of membership of other classes is also provided. This feature deals very well with cases that do not fit neatly into a larger class. Lastly, Autoclass requires the user to specify the measurement error of continuous variables.

Catchments were derived from the Australian digital elevation model. Physiographic data werederived from national spatial data sets. There was very little information on measurement errors for the spatial data, and so a conservative error of 5% of data range was adopted for all continuous attributes. The incorporation of uncertainty into spatial data sets remains a research challenge.

The results of the classification were very encouraging. The software found nine classes of catchments in the Murray Darling Basin. The classes grouped together geographically, and followed altitude and latitude gradients, despite the fact that these variables were not included in the classification. Descriptions of the classes reveal very different physiographic environments, ranging from dry and flat catchments (i.e. lowlands), through to wet and hilly catchments (i.e. mountainous areas). Rainfall and slope were two important discriminators between classes. These two attributes, in particular, will affect the ways in which the stream interacts with the catchment, and can thus be expected to modify the effects of land use change on ecological condition. Thus, realistic models of the effects of land use change on streams would differ between the different types of catchments, and sound management practices will differ.

A small number of catchments were assigned to their primary class with relatively low probability. These catchments lie on the boundaries of groups of catchments, with the second most likely class being an adjacent group. The locations of these ‘uncertain’ catchments show that the Bayesian classification dealt well with cases that do not fit neatly into larger classes.

Although the results are intuitive, we cannot yet assess whether the classifications described in this paper would assist the modelling of catchment scale effects on stream ecological condition. It is most likely that catchment classification and modelling will be an iterative process, where the needs of the model are used to guide classification, and the results of classifications used to suggest further refinements to models.

Classification of face images using local iterated function systems

There has been an increasing interest in face recognition in recent years. Many recognition methods have been developed so far, some very encouraging. A key remaining issue is the existence of variations in the input face image. Today, methods exist that can handle specific image variations. But we are yet to see methods that can be used more effectively in unconstrained situations. This paper presents a method that can handle partial translation, rotation, or scale variations in the input face image. The principal is to automatically identify objects within images using their partial self-similarities. The paper presents two recognition methods which can be used to recognise objects within images. A face recognition system is then presented that is insensitive to limited translation, rotation, or scale variations in the input face image. The performance of the system is evaluated through four experiments. The results show that the system achieves higher recognition rates than those of a number of existing approaches.

Detection and classification of road signs in natural environments

An automatic road sign recognition system first locates road signs within images captured by an imaging sensor on-board of a vehicle, and then identifies the detected road signs. This paper presents an automatic neural-network-based road sign recognition system. First, a study of the existing road sign recognition research is presented. In this study, the issues associated with automatic road sign recognition are described, the existing methods developed to tackle the road sign recognition problem are reviewed, and a comparison of the features of these methods is given. Second, the developed road sign recognition system is described. The system is capable of analysing live colour road scene images, detecting multiple road signs within each image, and classifying the type of road signs detected. The system consists of two modules: detection and classification. The detection module segments the input image in the hue-saturation-intensity colour space, and then detects road signs using a Multi-layer Perceptron neural-network. The classification module determines the type of detected road signs using a series of one to one architectural Multi-layer Perceptron neural networks. Two sets of classifiers are trained using the Resillient-Backpropagation and Scaled-Conjugate-Gradient algorithms. The two modules of the system are evaluated individually first. Then the system is tested as a whole. The experimental results demonstrate that the system is capable of achieving an average recognition hit-rate of 95.96% using the scaled-conjugate-gradient trained classifiers.

Determining clinical pharmacy workload by patient disease classification in medical and surgical patients

Aim: To determine the time needed to provide clinical pharmacy services to individual patient episodes for medical and surgical patients and the effect of patient presentation and complexity on the clinical pharmacy workload. Method: During a 5-month period in 2006 at two general hospitals, pharmacists recorded a defined range of activities that they provided for patients, including the actual times required for these tasks. A customised database linked to the two hospitals' patient administration systems stored the data according to the specific patient episode number. The influence of patient presentation and complexity on the clinical pharmacy activities provided was also examined. Results: The average time required by pharmacists to undertake a medication history interview and medication reconciliation was 9.6 (SD 4.9) minutes. Interventions required 5.7 (SD 4.6) minutes, clinical review of the medical record 5.5 (SD 4.0) minutes and medication order review 3.5 (SD 2.0) minutes. For all of these activities, the time required for medical patients was greater than for surgical patients and greater for 'complicated' patients. The average time required to perform all clinical pharmacy activities for 1071 completed patient episodes was 14.4 (SD 10.9) minutes and was greater for medical and 'complicated' patients. Conclusion: The time needed to provide clinical pharmacy services was affected by whether the patients were medical or surgical. The existence of comorbidities or complications affected these times. The times required to perform clinical pharmacy activities may not be consistent with recently proposed staff ratios for the provision of a basic clinical pharmacy service.

Function length as a tool for malware classification

The proliferation of malware is a serious threat to computer and information systems throughout the world. Antimalware companies are continually challenged to identify and counter new malware as it is released into the wild. In attempts to speed up this identification and response, many researchers have examined ways to efficiently automate classification of malware as it appears in the environment. In this paper, we present a fast, simple and scalable method of classifying Trojans based only on the lengths of their functions. Our results indicate that function length may play a significant role in classifying malware, and, combined with other features, may result in a fast, inexpensive and scalable method of malware classification.

An initial classification of branding strategy in Australian arts organisations

This research examined the corporate branding approaches and strategies adopted by six prominent Australian arts and cultural organisations. The aim of this exploration was to identify patterns in branding across different arts and cultural organisations, and attempt to provide an initial classification for understanding how these organisations approach branding strategy. We found that three factors influenced branding strategy in the surveyed organisations, viz., the focus of branding process, the degree of consistency in branding communication, and the required level of customers’ involvement in the branded products. The organisations studied were then plotted on a continuum that considered each of these factors.

A classification for content adaptation system

Content adaptation is an attractive solution for the ever growing desktop based Web content delivered to the user via heterogeneous devices, in order to provide acceptable experience while surfing the Web. Bridging the mismatch between the rich content and the user device's resources (display, processing, navigation, network bandwidth, media support) without user intervention requires a proactive behavior. While content adaptation poses multitude of benefits, without proper strategies, adaptation will not be truly optimized. There have been many projects focused on content adaptation that have been designed with different goals and approaches. In this paper, we introduce a comprehensive classification for content adaptation system. The classification is used to group the approaches applied in the implementation of existing content adaptation system. Survey on some content adaptation systems also been provided. We also present the research spectrum in content adaptation and discuss the challenges.

Using Choquet integrals for kNN approximation and classification

k-nearest neighbors (kNN) is a popular method for function approximation and classification. One drawback of this method is that the nearest neighbors can be all located on one side of the point in question x. An alternative natural neighbors method is expensive for more than three variables. In this paper we propose the use of the discrete Choquet integral for combining the values of the nearest neighbors so that redundant information is canceled out. We design a fuzzy measure based on location of the nearest neighbors, which favors neighbors located all around x.

Lung nodules detection by ensemble classification

A method is presented that achieves lung nodule detection by classification of nodule and non-nodule patterns. It is based on random forests which are ensemble learners that grow classification trees. Each tree produces a classification decision, and an integrated output is calculated. The performance of the developed method is compared against that of the support vector machine and the decision tree methods. Three experiments are performed using lung scans of 32 patients including thousands of images within which nodule locations are marked by expert radiologists. The classification errors and execution times are presented and discussed. The lowest classification error (2.4%) has been produced by the developed method.