Spatio temporal feature evaluation for action recognition

Spatio-Temporal interest points are the most popular feature representation in the field of action recognition. A variety of methods have been proposed to detect and describe local patches in video with several techniques reporting state of the art performance for action recognition. However, the reported results are obtained under different experimental settings with different datasets, making it difficult to compare the various approaches. As a result of this, we seek to comprehensively evaluate state of the art spatio- temporal features under a common evaluation framework with popular benchmark datasets (KTH, Weizmann) and more challenging datasets such as Hollywood2. The purpose of this work is to provide guidance for researchers, when selecting features for different applications with different environmental conditions. In this work we evaluate four popular descriptors (HOG, HOF, HOG/HOF, HOG3D) using a popular bag of visual features representation, and Support Vector Machines (SVM)for classification. Moreover, we provide an in-depth analysis of local feature descriptors and optimize the codebook sizes for different datasets with different descriptors. In this paper, we demonstrate that motion based features offer better performance than those that rely solely on spatial information, while features that combine both types of data are more consistent across a variety of conditions, but typically require a larger codebook for optimal performance.

Parking, parties and pets : disputes - the dark side of community living

In the last decade community living, in master planned communities or strata titled complexes, has increased. As land becomes scarcer, the popularity of these schemes is predicted to grow. Offsetting this popularity is the peculiarities of community living, in particular the often unthought-of difficulties arising from living in very close proximity to your neighbour. Such difficulties affect both amenity of life and property value. This paper seeks to inform practitioners of the issues arising from community living. It does this by identifying the more common forms of disputes and considering recent tribunal and court decisions. The paper concludes by identifying the dispute warning signs to assist to practitioners with the valuation process.

Spatial modelling of population health screening : the role of accessibility in predicting expected patient flows

This paper describes a generalised linear mixed model (GLMM) approach for understanding spatial patterns of participation in population health screening, in the presence of multiple screening facilities. The models presented have dual focus, namely the prediction of expected patient flows from regions to services and relative rates of participation by region- service combination, with both outputs having meaningful implications for the monitoring of current service uptake and provision. The novelty of this paper lies with the former focus, and an approach for distributing expected participation by region based on proximity to services is proposed. The modelling of relative rates of participation is achieved through the combination of different random effects, as a means of assigning excess participation to different sources. The methodology is applied to participation data collected from a government-funded mammography program in Brisbane, Australia.

On the commonly accepted assumptions regarding observed motor vehicle crash counts at transport system locations

Readily accepted knowledge regarding crash causation is consistently omitted from efforts to model and subsequently understand motor vehicle crash occurrence and their contributing factors. For instance, distracted and impaired driving accounts for a significant proportion of crash occurrence, yet is rarely modeled explicitly. In addition, spatially allocated influences such as local law enforcement efforts, proximity to bars and schools, and roadside chronic distractions (advertising, pedestrians, etc.) play a role in contributing to crash occurrence and yet are routinely absent from crash models. By and large, these well-established omitted effects are simply assumed to contribute to model error, with predominant focus on modeling the engineering and operational effects of transportation facilities (e.g. AADT, number of lanes, speed limits, width of lanes, etc.) The typical analytical approach—with a variety of statistical enhancements—has been to model crashes that occur at system locations as negative binomial (NB) distributed events that arise from a singular, underlying crash generating process. These models and their statistical kin dominate the literature; however, it is argued in this paper that these models fail to capture the underlying complexity of motor vehicle crash causes, and thus thwart deeper insights regarding crash causation and prevention. This paper first describes hypothetical scenarios that collectively illustrate why current models mislead highway safety researchers and engineers. It is argued that current model shortcomings are significant, and will lead to poor decision-making. Exploiting our current state of knowledge of crash causation, crash counts are postulated to arise from three processes: observed network features, unobserved spatial effects, and ‘apparent’ random influences that reflect largely behavioral influences of drivers. It is argued; furthermore, that these three processes in theory can be modeled separately to gain deeper insight into crash causes, and that the model represents a more realistic depiction of reality than the state of practice NB regression. An admittedly imperfect empirical model that mixes three independent crash occurrence processes is shown to outperform the classical NB model. The questioning of current modeling assumptions and implications of the latent mixture model to current practice are the most important contributions of this paper, with an initial but rather vulnerable attempt to model the latent mixtures as a secondary contribution.

Shaky Ground

Shaky Ground was a solo exhibition of works by Charles Robb held at Ryan Renshaw gallery, Brisbane in 2012. The exhibition comprised three sculptural works: a white rotating roundel with a drawing of the artist as seen from above; an artificial rock with a spinning aniseed ball nestled in one of its fissures; and a sculptural portrait of the artist dressed in a protective dust suit which was mounted perpendicular to the wall. The works were derivations or reorientations of previously exhibited work and established an ambiguous field of associations with each other based on formal characteristics or their proximity to the production site and processes. In so doing, the work formed part of the artist's ongoing exploration of sculpture, subjectivity and autogenous approaches to art practice.

Service condition of railroad corridors around the insulated rail joints

Railroad corridors contain large number of Insulated Rail Joints (IRJs) that act as safety critical elements in the circuitries of the signaling and broken rail identification systems. IRJs are regarded as sources of excitation for the passage of loaded wheels leading to high impact forces; these forces in turn cause dips, cross levels and twists to the railroad geometry in close proximity to the sections containing the IRJs in addition to the local damages to the railhead of the IRJs. Therefore, a systematic monitoring of the IRJs in railroad is prudent to mitigate potential risk of their sudden failure (e.g., broken tie plates) under the traffic. This paper presents a simple method of periodic recording of images using time-lapse photography and total station surveying measurements to understand the ongoing deterioration of the IRJs and their surroundings. Over a 500 day period, data were collected to examine the trends in narrowing of the joint gap due to plastic deformation the railhead edges and the dips, cross levels and twists caused to the railroad geometry due to the settlement of ties (sleepers) around the IRJs. The results reflect that the average progressive settlement beneath the IRJs is larger than that under the continuously welded rail, which leads to excessive deviation of railroad profile, cross levels and twists.

Vertical infrastructure inspection using a quadcopter and shared autonomy control

This paper presents a shared autonomy control scheme for a quadcopter that is suited for inspection of vertical infrastructure — tall man-made structures such as streetlights, electricity poles or the exterior surfaces of buildings. Current approaches to inspection of such structures is slow, expensive, and potentially hazardous. Low-cost aerial platforms with an ability to hover now have sufficient payload and endurance for this kind of task, but require significant human skill to fly. We develop a control architecture that enables synergy between the ground-based operator and the aerial inspection robot. An unskilled operator is assisted by onboard sensing and partial autonomy to safely fly the robot in close proximity to the structure. The operator uses their domain knowledge and problem solving skills to guide the robot in difficult to reach locations to inspect and assess the condition of the infrastructure. The operator commands the robot in a local task coordinate frame with limited degrees of freedom (DOF). For instance: up/down, left/right, toward/away with respect to the infrastructure. We therefore avoid problems of global mapping and navigation while providing an intuitive interface to the operator. We describe algorithms for pole detection, robot velocity estimation with respect to the pole, and position estimation in 3D space as well as the control algorithms and overall system architecture. We present initial results of shared autonomy of a quadrotor with respect to a vertical pole and robot performance is evaluated by comparing with motion capture data.

Decentralized detector generation in cooperative intrusion detection system

We consider Cooperative Intrusion Detection System (CIDS) which is a distributed AIS-based (Artificial Immune System) IDS where nodes collaborate over a peer-to-peer overlay network. The AIS uses the negative selection algorithm for the selection of detectors (e.g., vectors of features such as CPU utilization, memory usage and network activity). For better detection performance, selection of all possible detectors for a node is desirable but it may not be feasible due to storage and computational overheads. Limiting the number of detectors on the other hand comes with the danger of missing attacks. We present a scheme for the controlled and decentralized division of detector sets where each IDS is assigned to a region of the feature space. We investigate the trade-off between scalability and robustness of detector sets. We address the problem of self-organization in CIDS so that each node generates a distinct set of the detectors to maximize the coverage of the feature space while pairs of nodes exchange their detector sets to provide a controlled level of redundancy. Our contribution is twofold. First, we use Symmetric Balanced Incomplete Block Design, Generalized Quadrangles and Ramanujan Expander Graph based deterministic techniques from combinatorial design theory and graph theory to decide how many and which detectors are exchanged between which pair of IDS nodes. Second, we use a classical epidemic model (SIR model) to show how properties from deterministic techniques can help us to reduce the attack spread rate.

How do supercontinents break up? — assessing the continental large igneous province (LIP) record of the break-up of Pangea

LIP emplacement is linked to the timing and evolution of supercontinental break-up. LIP-related break-up produces volcanic rifted margins, new and large (up to 108 km2) ocean basins, and new, smaller continents that undergo dispersal and potentially reassembly (e.g., India). However, not all continental LIPs lead to continental rupture. We analysed the <330 Ma continental LIP record(following final assembly of Pangea) to find relationships between LIP event attributes (e.g., igneous volume, extent, distance from pre-existing continental margin) and ocean basin attributes (e.g., length of new ocean basin/rifted margin) and how these varied during the progressive break up of Pangea. No correlation exists between LIP magnitude and size of the subsequent ocean basin or rifted margin. Our review suggests a three-phased break-up history of Pangea: 1) “Preconditioning” phase (∼330–200 Ma): LIP events (n=7) occurred largely around the supercontinental margin clustering today in Asia, with a low (<20%) rifting success rate. The Panjal Traps at ∼280 Ma may represent the first continental rupturing event of Pangea, resulting in continental ribboning along the Tethyan margin; 2) “Main Break-up” phase (∼200–100 Ma): numerous large LIP events(n=10) in the supercontinent interior, resulting in highly successful fragmentation (90%) and large, new ocean basins(e.g., Central/South Atlantic, Indian, >3000 km long); 3) “Waning” phase (∼100–0 Ma): Declining LIP magnitudes (n=6), greater proximity to continental margins (e.g., Madagascar, North Atlantic, Afro-Arabia, Sierra Madre) producing smaller ocean basins (<2600 km long). How Pangea broke up may thus have implications for earlier supercontinent reconstructions and LIP record.

Distinguishing new science from calibration effects in the electron-volt neutron spectrometer VESUVIO at ISIS

The "standard" procedure for calibrating the Vesuvio eV neutron spectrometer at the ISIS neutron source, forming the basis for data analysis over at least the last decade, was recently documented in considerable detail by the instrument’s scientists. Additionally, we recently derived analytic expressions of the sensitivity of recoil peak positions with respect to fight-path parameters and presented neutron–proton scattering results that together called in to question the validity of the "standard" calibration. These investigations should contribute significantly to the assessment of the experimental results obtained with Vesuvio. Here we present new results of neutron–deuteron scattering from D2 in the backscattering angular range (theata > 90 degrees) which are accompanied by a striking energy increase that violates the Impulse Approximation, thus leading unequivocally the following dilemma: (A) either the "standard" calibration is correct and then the experimental results represent a novel quantum dynamical effect of D which stands in blatant contradiction of conventional theoretical expectations; (B) or the present "standard" calibration procedure is seriously deficient and leads to artificial outcomes. For Case(A), we allude to the topic of attosecond quantumdynamical phenomena and our recent neutron scattering experiments from H2 molecules. For Case(B),some suggestions as to how the "standard" calibration could be considerably improved are made.

Sensitivity of peak positions to flight-path parameters in a deep-inelastic scattering neutron TOF spectrometer

The effects of small changes in flight-path parameters (primary and secondary flight paths, detector angles), and of displacement of the sample along the beam axis away from its ideal position, are examined for an inelastic time-of-flight (TOF) neutron spectrometer, emphasising the deep-inelastic regime. The aim was to develop a rational basis for deciding what measured shifts in the positions of spectral peaks could be regarded as reliable in the light of the uncertainties in the calibrated flight-path parameters. Uncertainty in the length of the primary or secondary flight path has the least effect on the positions of the peaks of H, D and He, which are dominated by the accuracy of the calibration of the detector angles. This aspect of the calibration of a TOF spectrometer therefore demands close attention to achieve reliable outcomes where the position of the peaks is of significant scientific interest and is discussed in detail. The corresponding sensitivities of the position of peak of the Compton profile, J(y), to flight-path parameters and sample position are also examined, focusing on the comparability across experiments of results for H, D and He. We show that positioning the sample to within a few mm of the ideal position is required to ensure good comparability between experiments if data from detectors at high forward angles are to be reliably interpreted.

Algorithm for queue estimation with loop detector of time occupancy in off-ramps on signalized motorways

Long traffic queues on off-ramps significantly compromise the safety and throughput of motorways. Obtaining accurate queue information is crucial for countermeasure strategies. However, it is challenging to estimate traffic queues with locally installed inductive loop detectors. This paper deals with the problem of queue estimation with the interpretation of queuing dynamics and the corresponding time-occupancy distribution over motorway off-ramps. A novel algorithm for real-time queue estimation with two detectors is presented and discussed. Results derived from microscopic traffic simulation validated the effectiveness of the algorithm and revealed some of its useful features: (a) long and intermediate traffic queues could be accurately measured, (b) relatively simple detector input (i.e., time occupancy) was required, and (c) the estimation philosophy was independent with signal timing changes and provided the potential to cooperate with advanced strategies for signal control. Some issues concerning field implementation are also discussed.

Traffic queue estimation for metered motorway on-ramps through use of loop detector time occupancies

The primary objective of this study is to develop a robust queue estimation algorithm for motorway on-ramps. Real-time queue information is a vital input for dynamic queue management on metered on-ramps. Accurate and reliable queue information enables the management of on-ramp queue in an adaptive manner to the actual traffic queue size and thus minimises the adverse impacts of queue flush while increasing the benefit of ramp metering. The proposed algorithm is developed based on the Kalman filter framework. The fundamental conservation model is used to estimate the system state (queue size) with the flow-in and flow-out measurements. This projection results are updated with the measurement equation using the time occupancies from mid-link and link-entrance loop detectors. This study also proposes a novel single point correction method. This method resets the estimated system state to eliminate the counting errors that accumulate over time. In the performance evaluation, the proposed algorithm demonstrated accurate and reliable performances and consistently outperformed the benchmarked Single Occupancy Kalman filter (SOKF) method. The improvements over SOKF are 62% and 63% in average in terms of the estimation accuracy (MAE) and reliability (RMSE), respectively. The benefit of the innovative concepts of the algorithm is well justified by the improved estimation performance in congested ramp traffic conditions where long queues may significantly compromise the benchmark algorithm’s performance.

A Kalman filter based queue estimation algorithm using time occupancies for motorway on-ramps

The primary objective of this study is to develop a robust queue estimation algorithm for motorway on-ramps. Real-time queue information is the most vital input for a dynamic queue management that can treat long queues on metered on-ramps more sophistically. The proposed algorithm is developed based on the Kalman filter framework. The fundamental conservation model is used to estimate the system state (queue size) with the flow-in and flow-out measurements. This projection results are updated with the measurement equation using the time occupancies from mid-link and link-entrance loop detectors. This study also proposes a novel single point correction method. This method resets the estimated system state to eliminate the counting errors that accumulate over time. In the performance evaluation, the proposed algorithm demonstrated accurate and reliable performances and consistently outperformed the benchmarked Single Occupancy Kalman filter (SOKF) method. The improvements over SOKF are 62% and 63% in average in terms of the estimation accuracy (MAE) and reliability (RMSE), respectively. The benefit of the innovative concepts of the algorithm is well justified by the improved estimation performance in the congested ramp traffic conditions where long queues may significantly compromise the benchmark algorithm’s performance.