An evaluation of different features and learning models for anomalous event detection

The huge amount of CCTV footage available makes it very burdensome to process these videos manually through human operators. This has made automated processing of video footage through computer vision technologies necessary. During the past several years, there has been a large effort to detect abnormal activities through computer vision techniques. Typically, the problem is formulated as a novelty detection task where the system is trained on normal data and is required to detect events which do not fit the learned ‘normal’ model. There is no precise and exact definition for an abnormal activity; it is dependent on the context of the scene. Hence there is a requirement for different feature sets to detect different kinds of abnormal activities. In this work we evaluate the performance of different state of the art features to detect the presence of the abnormal objects in the scene. These include optical flow vectors to detect motion related anomalies, textures of optical flow and image textures to detect the presence of abnormal objects. These extracted features in different combinations are modeled using different state of the art models such as Gaussian mixture model(GMM) and Semi- 2D Hidden Markov model(HMM) to analyse the performances. Further we apply perspective normalization to the extracted features to compensate for perspective distortion due to the distance between the camera and objects of consideration. The proposed approach is evaluated using the publicly available UCSD datasets and we demonstrate improved performance compared to other state of the art methods.

Learning routines used by firms during construction of infrastructure projects using the alliance procurement method

Collaborative contracting has emerged over the past 15 years as an innovative project delivery framework that is particularly suited to infrastructure projects. Australia leads the world in the development of project and program alliance approaches to collaborative delivery. These approaches are considered to promise superior project results. However, very little is known about the learning routines that are most widely used in support of collaborative projects in general and alliance projects in particular. The literature on absorptive capacity and dynamic capabilities indicates that such learning enhances project performance. The learning routines employed at corporate level during the operation of collaborative infrastructure projects in Australia were examined through a large survey conducted in 2013. This paper presents a descriptive summary of the preliminary findings. The survey captured the experiences of 320 practitioners of collaborative construction projects, including public and private sector clients, contractors, consultants and suppliers (three per cent of projects were located in New Zealand, but for brevity’s sake the sample is referred to as Australian). The majority of projects identified used alliances (78.6%); whilst 9% used Early Contractor Involvement (ECI) contracts and 2.7% used Early Tender Involvement contracts, which are ‘slimmer’ types of collaborative contract. The remaining 9.7% of respondents used traditional contracts that employed some collaborative elements. The majority of projects were delivered for public sector clients (86.3%), and/or clients experienced with asset procurement (89.6%). All of the projects delivered infrastructure assets; one third in the road sector, one third in the water sector, one fifth in the rail sector, and the rest spread across energy, building and mining. Learning routines were explored within three interconnected phases: knowledge exploration, transformation and exploitation. The results show that explorative and exploitative learning routines were applied to a similar extent. Transformative routines were applied to a relatively low extent. It was also found that the most highly applied routine is ‘regularly applying new knowledge to collaborative projects’; and the least popular routine was ‘staff incentives to encourage information sharing about collaborative projects’. Future research planned by the authors will examine the impact of these routines on project performance.

Incorporating anchored learning in a C# Intelligent Tutoring System

Learning programming is known to be difficult. One possible reason why students fail programming is related to the fact that traditional learning in the classroom places more emphasis on lecturing the material instead of applying the material to a real application. For some students, this teaching model may not catch their interest. As a result they may not give their best effort to understand the material given. Seeing how the knowledge can be applied to real life problems can increase student interest in learning. As a consequence, this will increase their effort to learn. Anchored learning that applies knowledge to solve real life problems may be the key to improving student performance. In anchored learning, it is necessary to provide resources that can be accessed by the student as they learn. These resources can be provided by creating an Intelligent Tutoring System (ITS) that can support the student when they need help or experience a problem. Unfortunately, there is no ITS developed for the programming domain that has incorporated anchored learning in its teaching system. Having an ITS that supports anchored learning will not only be able to help the student learn programming effectively but will also make the learning process more enjoyable. This research tries to help students learn C# programming using an anchored learning ITS named CSTutor. Role playing is used in CSTutor to present a real world situation where they develop their skills. A knowledge base using First Order Logic is used to represent the student's code and to give feedback and assistance accordingly.

Pulse-coupled neural network performance for real-time identification of vegetation during forced landing

Safety concerns in the operation of autonomous aerial systems require safe-landing protocols be followed during situations where the mission should be aborted due to mechanical or other failure. This article presents a pulse-coupled neural network (PCNN) to assist in the vegetation classification in a vision-based landing site detection system for an unmanned aircraft. We propose a heterogeneous computing architecture and an OpenCL implementation of a PCNN feature generator. Its performance is compared across OpenCL kernels designed for CPU, GPU, and FPGA platforms. This comparison examines the compute times required for network convergence under a variety of images to determine the plausibility for real-time feature detection.

Using performance-based evaluation to close the loop between biological and robotic navigation

In this paper we describe the benefits of a performance-based approach to modeling biological systems for use in robotics. Specifically, we describe the RatSLAM system, a computational model of the navigation processes thought to drive navigation in a part of the rodent brain called the hippocampus. Unlike typical computational modeling approaches, which focus on biological fidelity, RatSLAM’s development cycle has been driven primarily by performance evaluation on robots navigating in a wide variety of challenging, real world environments. We briefly describe three seminal results, two in robotics and one in biology. In addition, we present current research on brain-inspired learning algorithms with the aim of enabling a robot to autonomously learn how best to use its sensor suite to navigate, without requiring any specific knowledge of the robot, sensor types or environment characteristics. Our aim is to drive discussion on the merits of practical, performance-focused implementations of biological models in robotics.

The role of the market sub-system and the socio-technical sub-system in innovation and firm performance : a dynamic capabilities approach

Organizational learning has been studied as a key factor in firm performance and internationalization. Moving beyond the past emphasis on market learning, we develop a more complete explanation of learning, its relationship to innovation, and their joint effect on early internationalization. We theorize that, driven by the founders’ international vision, early internationalizing firms employ a dual subsystem of dynamic capabilities: a market subsystem consisting of market-focused learning capability and marketing capability, and a socio-technical subsystem comprised of network learning capability and internally focused learning capability. We argue that innovation mediates the proposed relationship between the dynamic capability structure and early internationalization. We conduct case studies to develop the conceptual framework and test it in a field survey of early internationalizing firms from Australia and the United States. Our findings indicate a complex interplay of capabilities driving innovation and early internationalization. We provide theoretical and practical implications and offer insights for future research.

Improving megasite management performance through incentives : lessons learned from the Shanghai Expo construction

As the biggest expo site in history, construction of the Shanghai Expo site faced a lot of challenges, including involvement of lots of investors, megaconstruction scale, concurrent construction mode, involvement of more than 40,000 migrant workers, and extremely tight completion deadlines, among others. Consequently, these challenges imposed great obstacles on accomplishing the safety, quality, and environmental goals. Through a case study of the Shanghai Expo construction, this paper paper presents the design and implementation of multicriteria incentives in megaprojects to accomplish the safety, quality, and environmental goals. Both quantitative and qualitative findings were triangulated to demonstrate the outcome of the incentives. Six critical success factors (CSFs) for the incentives, rule design, process orientation, top management support, training and promotion, communication in process, and process learning and improvement are identified and validated through case study data and content analysis. It is believed that the findings of this paper can enhance understanding of multicriteria incentive schemes in general and provide insights in implementing these incentive schemes in future megaprojects, particularly in the People’s Republic of China (PRC).

Classification of cancer-related death certificates using machine learning

Background Cancer monitoring and prevention relies on the critical aspect of timely notification of cancer cases. However, the abstraction and classification of cancer from the free-text of pathology reports and other relevant documents, such as death certificates, exist as complex and time-consuming activities. Aims In this paper, approaches for the automatic detection of notifiable cancer cases as the cause of death from free-text death certificates supplied to Cancer Registries are investigated. Method A number of machine learning classifiers were studied. Features were extracted using natural language techniques and the Medtex toolkit. The numerous features encompassed stemmed words, bi-grams, and concepts from the SNOMED CT medical terminology. The baseline consisted of a keyword spotter using keywords extracted from the long description of ICD-10 cancer related codes. Results Death certificates with notifiable cancer listed as the cause of death can be effectively identified with the methods studied in this paper. A Support Vector Machine (SVM) classifier achieved best performance with an overall F-measure of 0.9866 when evaluated on a set of 5,000 free-text death certificates using the token stem feature set. The SNOMED CT concept plus token stem feature set reached the lowest variance (0.0032) and false negative rate (0.0297) while achieving an F-measure of 0.9864. The SVM classifier accounts for the first 18 of the top 40 evaluated runs, and entails the most robust classifier with a variance of 0.001141, half the variance of the other classifiers. Conclusion The selection of features significantly produced the most influences on the performance of the classifiers, although the type of classifier employed also affects performance. In contrast, the feature weighting schema created a negligible effect on performance. Specifically, it is found that stemmed tokens with or without SNOMED CT concepts create the most effective feature when combined with an SVM classifier.

Comparing different approach and avoidance models of learning and personality in the prediction of work, university, and leadership outcomes

Jackson (2005) developed a hybrid model of personality and learning, known as the learning styles profiler (LSP) which was designed to span biological, socio-cognitive, and experiential research foci of personality and learning research. The hybrid model argues that functional and dysfunctional learning outcomes can be best understood in terms of how cognitions and experiences control, discipline, and re-express the biologically based scale of sensation-seeking. In two studies with part-time workers undertaking tertiary education (N equals 137 and 58), established models of approach and avoidance from each of the three different research foci were compared with Jackson's hybrid model in their predictiveness of leadership, work, and university outcomes using self-report and supervisor ratings. Results showed that the hybrid model was generally optimal and, as hypothesized, that goal orientation was a mediator of sensation-seeking on outcomes (work performance, university performance, leader behaviours, and counterproductive work behaviour). Our studies suggest that the hybrid model has considerable promise as a predictor of work and educational outcomes as well as dysfunctional outcomes.

Hunting and gathering : new imperatives in mapping and collecting student learning data to assure quality outcomes

Assurance of learning (AOL) is a quality enhancement and quality assurance process used in higher education. It involves a process of determining programme learning outcomes and standards, and systematically gathering evidence to measure students' performance on these. The systematic assessment of whole-of-programme outcomes provides a basis for curriculum development and management, continuous improvement, and accreditation. To better understand how AOL processes operate, a national study of university practices across one discipline area, business and management, was undertaken. To solicit data on AOL practice, interviews were undertaken with a sample of business school representatives (n = 25). Two key processes emerged: (1) mapping of graduate attributes and (2) collection of assurance data. External drivers such as professional accreditation and government legislation were the primary reasons for undertaking AOL outcomes but intrinsic motivators in relation to continuous improvement were also evident. The facilitation of academic commitment was achieved through an embedded approach to AOL by the majority of universities in the study. A sustainable and inclusive process of AOL was seen to support wider stakeholder engagement in the development of higher education learning outcomes.

Translations of industry-based student learning and academic performances

We hypothesized that Industry based learning and teaching, especially through industry assigned student projects or training programs, is an integral part of science, technology, engineering and mathematics (STEM) education. In this paper we show that industry-based student training and experience increases students’ academic performances independent to the organizational parameters and contexts. The literature on industry-based student training focuses on employability and the industry dimension, and neglects in many ways the academic dimension. We observed that the association factors between academic attributes and contributions of industry-based student training are central and vital to the technological learning experiences. We explore international initiatives and statistics collected of student projects in two categories: Industry based learning performances and on campus performances. The data collected were correlated to five (5) universities in different industrialized countries, e.g., Australia N=545, Norway N=279, Germany N=74, France N=107 and Spain N=802 respectively. We analyzed industry-based student training along with company assigned student projects compared with in comparisons to campus performance. The data that suggests a strong correlation between industry-based student training per se and improved performance profiles or increasing motivation shows that industry-based student training increases student academic performance independent of organizational parameters and contexts. The programs we augmented were orthogonal to each other however, the trend of the students’ academic performances are identical. An isolated cohort for the reported countries that opposed our hypothesis warrants further investigation.

Peripheral vision benefits spatial learning by guiding eye movements

The loss of peripheral vision impairs spatial learning and navigation. However, the mechanisms underlying these impairments remain poorly understood. One advantage of having peripheral vision is that objects in an environment are easily detected and readily foveated via eye movements. The present study examined this potential benefit of peripheral vision by investigating whether competent performance in spatial learning requires effective eye movements. In Experiment 1, participants learned room-sized spatial layouts with or without restriction on direct eye movements to objects. Eye movements were restricted by having participants view the objects through small apertures in front of their eyes. Results showed that impeding effective eye movements made subsequent retrieval of spatial memory slower and less accurate. The small apertures also occluded much of the environmental surroundings, but the importance of this kind of occlusion was ruled out in Experiment 2 by showing that participants exhibited intact learning of the same spatial layouts when luminescent objects were viewed in an otherwise dark room. Together, these findings suggest that one of the roles of peripheral vision in spatial learning is to guide eye movements, highlighting the importance of spatial information derived from eye movements for learning environmental layouts.

Sequential versus simultaneous viewing of an environment : effects of focal attention to individual object locations on visual spatial learning

We investigated memories of room-sized spatial layouts learned by sequentially or simultaneously viewing objects from a stationary position. In three experiments, sequential viewing (one or two objects at a time) yielded subsequent memory performance that was equivalent or superior to simultaneous viewing of all objects, even though sequential viewing lacked direct access to the entire layout. This finding was replicated by replacing sequential viewing with directed viewing in which all objects were presented simultaneously and participants’ attention was externally focused on each object sequentially, indicating that the advantage of sequential viewing over simultaneous viewing may have originated from focal attention to individual object locations. These results suggest that memory representation of object-to-object relations can be constructed efficiently by encoding each object location separately, when those locations are defined within a single spatial reference system. These findings highlight the importance of considering object presentation procedures when studying spatial learning mechanisms.

Path information effects in visual and proprioceptive spatial learning

Objects in an environment are often encountered sequentially during spatial learning, forming a path along which object locations are experienced. The present study investigated the effect of spatial information conveyed through the path in visual and proprioceptive learning of a room-sized spatial layout, exploring whether different modalities differentially depend on the integrity of the path. Learning object locations along a coherent path was compared with learning them in a spatially random manner. Path integrity had little effect on visual learning, whereas learning with the coherent path produced better memory performance than random order learning for proprioceptive learning. These results suggest that path information has differential effects in visual and proprioceptive spatial learning, perhaps due to a difference in the way one establishes a reference frame for representing relative locations of objects.

Brain decoding based on functional magnetic resonance imaging using machine learning : a comparative study

Brain decoding of functional Magnetic Resonance Imaging data is a pattern analysis task that links brain activity patterns to the experimental conditions. Classifiers predict the neural states from the spatial and temporal pattern of brain activity extracted from multiple voxels in the functional images in a certain period of time. The prediction results offer insight into the nature of neural representations and cognitive mechanisms and the classification accuracy determines our confidence in understanding the relationship between brain activity and stimuli. In this paper, we compared the efficacy of three machine learning algorithms: neural network, support vector machines, and conditional random field to decode the visual stimuli or neural cognitive states from functional Magnetic Resonance data. Leave-one-out cross validation was performed to quantify the generalization accuracy of each algorithm on unseen data. The results indicated support vector machine and conditional random field have comparable performance and the potential of the latter is worthy of further investigation.