Improved adaptivity and robustness in decentralised multi-camera networks

In this paper we present increased adaptivity and robustness in distributed object tracking by multi-camera networks using a socio-economic mechanism for learning the vision graph. To build-up the vision graph autonomously within a distributed smart-camera network, we use an ant-colony inspired mechanism, which exchanges responsibility for tracking objects using Vickrey auctions. Employing the learnt vision graph allows the system to optimise its communication continuously. Since distributed smart camera networks are prone to uncertainties in individual cameras, such as failures or changes in extrinsic parameters, the vision graph should be sufficiently robust and adaptable during runtime to enable seamless tracking and optimised communication. To better reflect real smart-camera platforms and networks, we consider that communication and handover are not instantaneous, and that cameras may be added, removed or their properties changed during runtime. Using our dynamic socio-economic approach, the network is able to continue tracking objects well, despite all these uncertainties, and in some cases even with improved performance. This demonstrates the adaptivity and robustness of our approach.

CamSim:a distributed smart camera network simulator

Smart cameras allow pre-processing of video data on the camera instead of sending it to a remote server for further analysis. Having a network of smart cameras allows various vision tasks to be processed in a distributed fashion. While cameras may have different tasks, we concentrate on distributed tracking in smart camera networks. This application introduces various highly interesting problems. Firstly, how can conflicting goals be satisfied such as cameras in the network try to track objects while also trying to keep communication overhead low? Secondly, how can cameras in the network self adapt in response to the behavior of objects and changes in scenarios, to ensure continued efficient performance? Thirdly, how can cameras organise themselves to improve the overall network's performance and efficiency? This paper presents a simulation environment, called CamSim, allowing distributed self-adaptation and self-organisation algorithms to be tested, without setting up a physical smart camera network. The simulation tool is written in Java and hence allows high portability between different operating systems. Relaxing various problems of computer vision and network communication enables a focus on implementing and testing new self-adaptation and self-organisation algorithms for cameras to use.

Learning to be different:heterogeneity and efficiency in distributed smart camera networks

In this paper we study the self-organising behaviour of smart camera networks which use market-based handover of object tracking responsibilities to achieve an efficient allocation of objects to cameras. Specifically, we compare previously known homogeneous configurations, when all cameras use the same marketing strategy, with heterogeneous configurations, when each camera makes use of its own, possibly different marketing strategy. Our first contribution is to establish that such heterogeneity of marketing strategies can lead to system wide outcomes which are Pareto superior when compared to those possible in homogeneous configurations. However, since the particular configuration required to lead to Pareto efficiency in a given scenario will not be known in advance, our second contribution is to show how online learning of marketing strategies at the individual camera level can lead to high performing heterogeneous configurations from the system point of view, extending the Pareto front when compared to the homogeneous case. Our third contribution is to show that in many cases, the dynamic behaviour resulting from online learning leads to global outcomes which extend the Pareto front even when compared to static heterogeneous configurations. Our evaluation considers results obtained from an open source simulation package as well as data from a network of real cameras. © 2013 IEEE.

Comparison of subjective and objective methods to determine the retinal arterio-venous ratio using fundus photography

Purpose: To assess the inter and intra observer variability of subjective grading of the retinal arterio-venous ratio (AVR) using a visual grading and to compare the subjectively derived grades to an objective method using a semi-automated computer program. Methods: Following intraocular pressure and blood pressure measurements all subjects underwent dilated fundus photography. 86 monochromatic retinal images with the optic nerve head centred (52 healthy volunteers) were obtained using a Zeiss FF450+ fundus camera. Arterio-venous ratios (AVR), central retinal artery equivalent (CRAE) and central retinal vein equivalent (CRVE) were calculated on three separate occasions by one single observer semi-automatically using the software VesselMap (ImedosSystems, Jena, Germany). Following the automated grading, three examiners graded the AVR visually on three separate occasions in order to assess their agreement. Results: Reproducibility of the semi-automatic parameters was excellent (ICCs: 0.97 (CRAE); 0.985 (CRVE) and 0.952 (AVR)). However, visual grading of AVR showed inter grader differences as well as discrepancies between subjectively derived and objectively calculated AVR (all p < 0.000001). Conclusion: Grader education and experience leads to inter-grader differences but more importantly, subjective grading is not capable to pick up subtle differences across healthy individuals and does not represent true AVR when compared with an objective assessment method. Technology advancements mean we no longer rely on opthalmoscopic evaluation but can capture and store fundus images with retinal cameras, enabling us to measure vessel calibre more accurately compared to visual estimation; hence it should be integrated in optometric practise for improved accuracy and reliability of clinical assessments of retinal vessel calibres. © 2014 Spanish General Council of Optometry.

Self-organising zooms for decentralised redundancy management in visual sensor networks

When visual sensor networks are composed of cameras which can adjust the zoom factor of their own lens, one must determine the optimal zoom levels for the cameras, for a given task. This gives rise to an important trade-off between the overlap of the different cameras’ fields of view, providing redundancy, and image quality. In an object tracking task, having multiple cameras observe the same area allows for quicker recovery, when a camera fails. In contrast having narrow zooms allow for a higher pixel count on regions of interest, leading to increased tracking confidence. In this paper we propose an approach for the self-organisation of redundancy in a distributed visual sensor network, based on decentralised multi-objective online learning using only local information to approximate the global state. We explore the impact of different zoom levels on these trade-offs, when tasking omnidirectional cameras, having perfect 360-degree view, with keeping track of a varying number of moving objects. We further show how employing decentralised reinforcement learning enables zoom configurations to be achieved dynamically at runtime according to an operator’s preference for maximising either the proportion of objects tracked, confidence associated with tracking, or redundancy in expectation of camera failure. We show that explicitly taking account of the level of overlap, even based only on local knowledge, improves resilience when cameras fail. Our results illustrate the trade-off between maintaining high confidence and object coverage, and maintaining redundancy, in anticipation of future failure. Our approach provides a fully tunable decentralised method for the self-organisation of redundancy in a changing environment, according to an operator’s preferences.

Static, dynamic, and adaptive heterogeneity in distributed smart camera networks

We study heterogeneity among nodes in self-organizing smart camera networks, which use strategies based on social and economic knowledge to target communication activity efficiently. We compare homogeneous configurations, when cameras use the same strategy, with heterogeneous configurations, when cameras use different strategies. Our first contribution is to establish that static heterogeneity leads to new outcomes that are more efficient than those possible with homogeneity. Next, two forms of dynamic heterogeneity are investigated: nonadaptive mixed strategies and adaptive strategies, which learn online. Our second contribution is to show that mixed strategies offer Pareto efficiency consistently comparable with the most efficient static heterogeneous configurations. Since the particular configuration required for high Pareto efficiency in a scenario will not be known in advance, our third contribution is to show how decentralized online learning can lead to more efficient outcomes than the homogeneous case. In some cases, outcomes from online learning were more efficient than all other evaluated configuration types. Our fourth contribution is to show that online learning typically leads to outcomes more evenly spread over the objective space. Our results provide insight into the relationship between static, dynamic, and adaptive heterogeneity, suggesting that all have a key role in achieving efficient self-organization.

Self-aware and self-expressive camera networks

Smart cameras perform on-board image analysis, adapt their algorithms to changes in their environment, and collaborate with other networked cameras to analyze the dynamic behavior of objects. A proposed computational framework adopts the concepts of self-awareness and self-expression to more efficiently manage the complex tradeoffs among performance, flexibility, resources, and reliability. The Web extra at http://youtu.be/NKe31-OKLz4 is a video demonstrating CamSim, a smart camera simulation tool, enables users to test self-adaptive and self-organizing smart-camera techniques without deploying a smart-camera network.

Mobility changes in older age: neuropsychological, neurophysiological and cognitive predictors of successful adaptation in a real world scenario

The aims of this thesis were to investigate the neuropsychological, neurophysiological, and cognitive contributors to mobility changes with increasing age. In a series of studies with adults aged 45-88 years, unsafe pedestrian behaviour and falls were investigated in relation to i) cognitive functions (including response time variability, executive function, and visual attention tests), ii) mobility assessments (including gait and balance and using motion capture cameras), iii) motor initiation and pedestrian road crossing behavior (using a simulated pedestrian road scene), iv) neuronal and functional brain changes (using a computer based crossing task with magnetoencephalography), and v) quality of life questionnaires (including fear of falling and restricted range of travel). Older adults are more likely to be fatally injured at the far-side of the road compared to the near-side of the road, however, the underlying mobility and cognitive processes related to lane-specific (i.e. near-side or far-side) pedestrian crossing errors in older adults is currently unknown. The first study explored cognitive, motor initiation, and mobility predictors of unsafe pedestrian crossing behaviours. The purpose of the first study (Chapter 2) was to determine whether collisions at the near-side and far-side would be differentially predicted by mobility indices (such as walking speed and postural sway), motor initiation, and cognitive function (including spatial planning, visual attention, and within participant variability) with increasing age. The results suggest that near-side unsafe pedestrian crossing errors are related to processing speed, whereas far-side errors are related to spatial planning difficulties. Both near-side and far-side crossing errors were related to walking speed and motor initiation measures (specifically motor initiation variability). The salient mobility predictors of unsafe pedestrian crossings determined in the above study were examined in Chapter 3 in conjunction with the presence of a history of falls. The purpose of this study was to determine the extent to which walking speed (indicated as a salient predictor of unsafe crossings and start-up delay in Chapter 2), and previous falls can be predicted and explained by age-related changes in mobility and cognitive function changes (specifically within participant variability and spatial ability). 53.2% of walking speed variance was found to be predicted by self-rated mobility score, sit-to-stand time, motor initiation, and within participant variability. Although a significant model was not found to predict fall history variance, postural sway and attentional set shifting ability was found to be strongly related to the occurrence of falls within the last year. Next in Chapter 4, unsafe pedestrian crossing behaviour and pedestrian predictors (both mobility and cognitive measures) from Chapter 2 were explored in terms of increasing hemispheric laterality of attentional functions and inter-hemispheric oscillatory beta power changes associated with increasing age. Elevated beta (15-35 Hz) power in the motor cortex prior to movement, and reduced beta power post-movement has been linked to age-related changes in mobility. In addition, increasing recruitment of both hemispheres has been shown to occur and be beneficial to perform similarly to younger adults in cognitive tasks (Cabeza, Anderson, Locantore, & McIntosh, 2002). It has been hypothesised that changes in hemispheric neural beta power may explain the presence of more pedestrian errors at the farside of the road in older adults. The purpose of the study was to determine whether changes in age-related cortical oscillatory beta power and hemispheric laterality are linked to unsafe pedestrian behaviour in older adults. Results indicated that pedestrian errors at the near-side are linked to hemispheric bilateralisation, and neural overcompensation post-movement, 4 whereas far-side unsafe errors are linked to not employing neural compensation methods (hemispheric bilateralisation). Finally, in Chapter 5, fear of falling, life space mobility, and quality of life in old age were examined to determine their relationships with cognition, mobility (including fall history and pedestrian behaviour), and motor initiation. In addition to death and injury, mobility decline (such as pedestrian errors in Chapter 2, and falls in Chapter 3) and cognition can negatively affect quality of life and result in activity avoidance. Further, number of falls in Chapter 3 was not significantly linked to mobility and cognition alone, and may be further explained by a fear of falling. The objective of the above study (Study 2, Chapter 3) was to determine the role of mobility and cognition on fear of falling and life space mobility, and the impact on quality of life measures. Results indicated that missing safe pedestrian crossing gaps (potentially indicating crossing anxiety) and mobility decline were consistent predictors of fear of falling, reduced life space mobility, and quality of life variance. Social community (total number of close family and friends) was also linked to life space mobility and quality of life. Lower cognitive functions (particularly processing speed and reaction time) were found to predict variance in fear of falling and quality of life in old age. Overall, the findings indicated that mobility decline (particularly walking speed or walking difficulty), processing speed, and intra-individual variability in attention (including motor initiation variability) are salient predictors of participant safety (mainly pedestrian crossing errors) and wellbeing with increasing age. More research is required to produce a significant model to explain the number of falls.