It is hard to estimate optical flow given a realworld video sequence with camera shake and other motion blur. In this paper, we first investigate the blur parameterization for video footage using near linear motion elements. We then combine a commercial 3D pose sensor with an RGB camera, in order to film video footage of interest together with the camera motion. We illustrates that this additional camera motion/trajectory channel can be embedded into a hybrid framework by interleaving an iterative blind deconvolution and warping based optical flow scheme. Our method yields improved accuracy within three other state-of-the-art baselines given our proposed ground truth blurry sequences; and several other realworld sequences filmed by our imaging system.
Abstract. Normative systems offer a means to govern agent behaviour in dynamic open environments. Under the governance, agents themselves must be able to reason about compliance with state-or event-based norms (or both) depending upon the formalism used. This paper describes how norm awareness enables a BDI agent to exhibit norm compliant behaviour at run-time taking into account normative factors. To this end, we propose N-Jason, a run-time norm compliant BDI agent framework supporting norm-aware deliberation as well as run-time norm execution mechanism, through which new unknown norms are recognised and bring about the triggering of plans. To be able to process a norm such as an obligation, the agent architecture must be able to deal with deadlines and priorities, and choose between plans triggered by a particular norm. Consequently, we extend the syntax and the scheduling algorithm of AgentSpeak(RT) to operate in the context of Jason/AgentSpeak(L) and provide 'real-time agency', which we explain through a detailed examination of the operational semantics of a single reasoning cycle.
The use of polite agents is a new approach in order to improve efficiency and naturalism in navigation for player characters in crowded virtual worlds. This paper aims to model the politeness of virtual humans using logic-based approaches, subject to theory of politeness decomposed of conventional and interpersonal politeness. To do so, we propose a high-level agent architecture combined with normative framework to model and reason about 'polite' behaviours in social situations. With this architecture, we demonstrate (i) specifying polite behaviours as a form of social norms; (ii) generating polite behaviours using social reasoning technique; (iii) deliberation with such norms in belief-desire-intention agents; and (iv) realising physical actions based on the decision. Implementation for social reasoning is achieved by InstAL, based on the semantics of answer set programming. Using experiments with simple collision avoidance model, we show the effectiveness of polite behaviour in navigation designed by such architecture, as well as the adequacy of this architecture for modelling theory of politeness in all circumstances.
The conflicting evidence in the literature on energy feedback as a driver for energy behaviour change has lead to the realization that it is a complex problem and that interventions must be proposed and evaluated in the context of a tangled web of individual and societal factors. We put forward an integrated agent-based computational model of energy consumption behaviour change interventions based on personal values and energy literacy, informed by research in persuasive technologies, environmental, educational and cognitive psychology, sociology, and energy education. Our objectives are: (i) to build a framework to accommodate a rich variety of models that might impact consumption decisions, (ii) to use the simulation as a means to evaluate persuasive technologies in-silico prior to deployment. The model novelty lies in its capacity to connect the determinants of energy related behaviour (values, energy literacy and social practices) and several generic design strategies proposed in the area of persuasive technologies within one framework. We validate the framework using survey data and personal value and energy consumption data extracted from a 2-year field study in Exeter, UK. The preliminary evaluation results demonstrate that the model can predict energy saving behaviour much better than a random model and can correctly estimate the effect of persuasive technologies. The model can be embedded into an adaptive decision-making system for energy behaviour change. B Nataliya Mogles
Abstract. The development of and accessibility to rich virtual environments, both for recreation and training activities leads to the use of intelligent agents to control avatars (and other entities) in these environments. There is a fundamental tension in such systems between tight integration, for performance and low coupling, for generality, flexibility and extensibility. This paper addresses the engineering issues in connecting agent platforms and other software entities with virtual environments, driven by the following informal requirements: (i) accessibility: we would like (easily) to be able to connect any (legacy) software component with the virtual environment (ii) performance: we want the benefits of decoupling, but not at a high price in performance (iii) distribution: we would like to be able to locate functionality where needed, when necessary, but also be location agnostic otherwise (iv) scalability: we would like to support large-scale and geographically dispersed virtual environments. We start from the position that the basic currency unit of such systems can be events. We describe the Bath Sensor Framework, which is a middleware that attempts to satisfy the above goals and to provide a low-latency linking mechanism between event producers and event consumers, while minimising the effect of coupling of components. We illustrate the framework in two complementary case studies using the Jason agent platform, Second Life and AGAVE (a 3D VE for vehicles). Through these examples, we are able to carry out a preliminary evaluation of the approach against the factors above, against alternative systems and demonstrate effective distributed execution.
Although the latest energy-efficient buildings use a large number of sensors and measuring instruments to predict consumption more accurately, it is generally not possible to identify which data are the most valuable or key for analysis among the tens of thousands of data points. This study selected the electric energy as a subset of total building energy consumption because it accounts for more than 65% of the total building energy consumption, and identified the variables that contribute to electric energy use. However, this study aimed to confirm data from a building using clustering in machine learning, instead of a calculation method from engineering simulation, to examine the variables that were identified and determine whether these variables had a strong correlation with energy consumption. Three different methods confirmed that the major variables related to electric energy consumption were significant. This research has significance because it was able to identify the factors in electric energy, accounting for more than half of the total building energy consumption, that had a major effect on energy consumption and revealed that these key variables alone, not the default values of many different items in simulation analysis, can ensure the reliable prediction of energy consumption.
There is a growing desire to measure the operational performance of buildings-often many buildings simultaneously-but the cost of sensors and complexity of deployment is a significant constraint. In this paper, we present an approach to minimising the cost of sensing by recognising that researchers are often not interested in the raw data itself but rather some inferred performance metric (e.g. high CO 2 levels may indicate poor
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.