Search

A number of pro-superpixel based saliency models have recently been proposed, which segment the image into small perceptually homogeneous regions before saliency computation. Such approaches ignore important object properties, resulting in inappropriate object annotations and considerably different saliency assignment to the various regions of an object. Although previous techniques employ multi-scale saliency maps in an attempt to rectify this problem, it becomes difficult to retain the characteristics of proto-objects after the first stage of processing. We introduce matting components based saliency to address the problems of inappropriate object annotations and inappropriate saliency assignment to object regions. The matting components account for proto-object properties by employing object aware spectral segmentation. To complement the matting component based saliency, we also employ the smallest eigenvectors of a matting Laplacian matrix. Color spatial distribution features are employed to capture global relationships at the pixel-level and assist the process of matting components based saliency computation. A novel joint optimization framework is introduced to fuse the features and learn important associated parameters. The contributions of the proposed approach are two-fold. The first contribution is the introduction of proto-objects aware spectral segmentation to obtain an accurate foreground saliency. The second contribution is the joint optimization of important parameters in conjunction with learning feature importance. In contrast to superpixel based approaches, the proposed model is able to completely annotate salient objects and assign similar saliency to various regions of the salient object. Moreover, the proposed approach shows robust and efficient performance across five challenging benchmark datasets when compared with 10 recently proposed state-of-the-art saliency detection models. [ABSTRACT FROM AUTHOR]

Figure-ground segmentation is the process of separating regions of interest from unimportant background. One challenge is to segment images with high variations (e.g. containing a cluttered background), which requires effective feature sets to capture the distinguishing information between objects and backgrounds. Feature selection is necessary to remove noisy/redundant features from those extracted by image descriptors. As a powerful search algorithm, genetic programming (GP) is employed for the first time to build feature selection methods that aims to improve the segmentation performance of standard classification techniques. Both single-objective and multi-objective GP techniques are investigated, based on which three novel feature selection methods are proposed. Specifically, one method is single-objective, called PGP-FS (parsimony GP feature selection); while the other two are multi-objective, named nondominated sorting GP feature selection (NSGP-FS) and strength Pareto GP feature selection (SPGP-FS). The feature subsets produced by the three proposed methods, two standard sequential selection algorithms, and the original feature set are tested via standard classification algorithms on two datasets with high variations (the Weizmann and Pascal datasets). The results show that the two multi-objective methods (NSGP-FS and SPGP-FS) can produce feature subsets that lead to solutions achieving better segmentation performance with lower numbers of features than the sequential algorithms and the original feature set based on standard classifiers for given segmentation tasks. In contrast, PGP-FS produces results that are not consistent for different classifiers. This indicates that the proposed multi-objective methods can help standard classifiers improve the segmentation performance while reducing the processing time. Moreover, compared with SPGP-FS, NSGP-FS is equally capable of producing effective feature subsets, yet is better at keeping diverse solutions. [ABSTRACT FROM AUTHOR]

Salient object detection is the task of automatically localizing objects of interests in a scene by suppressing the background information, which facilitates various machine vision applications such as object segmentation, recognition and tracking. Combining features from different feature-modalities has been demonstrated to enhance the performance of saliency prediction algorithms and different feature combinations are often suited to different types of images. However, existing saliency learning techniques attempt to apply a single feature combination across all image types and thus lose generalization in the test phase when considering unseen images. Learning classifier systems (LCSs) are an evolutionary machine learning technique that evolve a set of rules, based on a niched genetic reproduction, which collectively solve the problem. It is hypothesized that the LCS technique has the ability to autonomously learn different feature combinations for different image types. Hence, this paper further investigates the application of LCS for learning image dependent feature fusion strategies for the task of salient object detection. The obtained results show that the proposed method outperforms, through evolving generalized rules to compute saliency maps, the individual feature based methods and seven combinatorial techniques in detecting salient objects from three well known benchmark datasets of various types and difficulty levels. [ABSTRACT FROM AUTHOR]

The parametric uncertainties inherent in the models of renewable and sustainable energy systems (RSESs) make the associated decision-making processes of integrated resource operation, planning, and designing profoundly complex. Accordingly, intelligent energy management strategies are recognised as an effective intervention to efficiently accommodate the variability inherent in various input data and integrate distributed demand-side flexibility resources. To identify the key methodological and content gaps in the area of stochastic dispatch and planning optimisation of RSESs in the presence of responsive loads, this paper systematically reviews and synthetically analyses 252 relevant peer-reviewed academic articles. The review reveals that academic studies have utilised a wide variety of methods for the joint quantification of uncertainties and procurement of demand response services, while optimally designing and scheduling RSESs. However, to minimise simulation-to-reality gaps, research is needed into more integrated energy optimisation models that simultaneously characterise a broader spectrum of problem-inherent uncertainties and make behaviourally-founded use of flexible demand-side resources. More specifically, the review finds that while the research in this area is rich in thematic scope, it is commonly associated with strong simplifying assumptions that disconnect the corresponding approaches from reality, and thereby obscure the real challenges of transferring simulations into the real world. Accordingly, based on the descriptive analyses conducted and knowledge gaps identified, the paper provides useful insights into myriad possibilities for new research to more effectively utilise the potential of responsive loads, whilst simultaneously characterising the most salient problem-inherent parametric sources of uncertainty, during the investment planning and operational phases of RSESs. [Display omitted] • A systematic review of stochastic energy optimisation with demand response is conducted. • 252 relevant publications are identified in the literature and synthetically analysed. • Relevant research is rich in thematic scope, but involves strong simplifying assumptions. • Challenges ahead and opportunities for future research in the area are discussed. • Computational burden remains the most significant barrier to minimising reality gaps. [ABSTRACT FROM AUTHOR]

• A new aggregator demand response model is shown to increase the flexibility of microgrids. • Strategic game theory is used to ensure a cost-optimal solution for sustainable energy systems. • Imported and dispatched distributed energy is optimised from the utility perspective. • This research provides new insight for assessing pre-feasibility and business case assessments. • Optimised system costs are reduced by up to 66% for the case of Ohakune, New Zealand. Aggregator-activated demand response (DR) is widely recognised as a viable means for increasing the flexibility of renewable and sustainable energy systems (RSESs) necessary to accommodate a high penetration of variable renewables. To this end, by acting as DR aggregators and offering energy trading capabilities to smaller customers, energy retailers unlock additional sources of demand-side flexibility to ensure the cost-optimal operation of RSESs. Accordingly, a growing body of literature has highlighted the ways in which non-cooperative game theory could be used to reduce the gaps between modelled and real-world results for aggregator-mediated DR schemes. This paper aims to contribute to the trends of giving a realistic grounding to research on distributed DR-integrated energy scheduling by using insights from non-cooperative game theory to determine: (1) the optimal trade-off between importing electricity and utilising DR capacity in grid-tied RSESs, (2) the impact of the price elasticity of DR supply of different customer classes – especially, new sources of electricity demand, such as e-mobility – on the system-level dispatch of DR resources, and (3) the financial implications of harnessing the flexibility potential of a large number of end-consumers across different sectors. Accordingly, the principal goal of the paper is to develop an operational planning optimisation model that can be directly applied to real-world aggregator-mediated, market-based demand-side flexibility provisioning domains. To this end, this paper presents the first DR elasticity-aware, non-cooperative game-theoretic DR scheduling model that: (1) yields the best compromise solution between imported power and dispatched DR resources from the utility's perspective, (2) characterises the utility-aggregator-customer interactions during the market-based DR trade process with several customer categories involved, and (3) disaggregates the total sectoral load on the system to individual end-consumers, which has potential implications for pre-feasibility and business case assessments. The application of the model to a conceptual micro-grid for the town of Ohakune, in New Zealand, demonstrates its effectiveness in reducing the daily system operational cost (over the critical peak hours) by ~66% and ~47% on a representative summer and winter day, respectively. Importantly, the paper provides statistically significant evidence supporting that activating the flexibility potential of small- to medium-scale end-consumers through specifically defined third-party aggregators in a market-based approach – that is aware of strategic interactions among instrumentally rational economic agents involved in the dispatch and delivery of DR resources – plays a significant role in the cost-optimal transition to 100%-renewable electricity generation systems within the smart grid paradigm. [ABSTRACT FROM AUTHOR]

• A market-driven model is devised for long-term projections of incentive-aware loads. • Responsive loads are integrated through dedicated aggregators for improved accuracy. • A level playing field is provided for fuel cell electric vehicle-to-grid technology. • An energy filter-based approach is employed to allocate various storage technologies. • The model's potential in cutting a test micro-grid's lifetime costs by 21% is shown. While industrial demand response programmes have long been valued to support the power grid, recent advances in information and communications technology have enabled new opportunities to leverage the potential of responsive loads in less energy-dense end-use sectors. This brings to light the importance of accurately projecting flexible demand-side resources in the long-term investment planning process of micro-grids. This paper introduces a customer comfort-aware, demand response-integrated long-term micro-grid planning optimisation model. The model (1) draws on non-cooperative game theory and the Stackelberg leadership principles to understand and reflect the strategic behaviour of energy utilities, demand response aggregators, and end-consumers, (2) produces optimal trade-offs between power imported from the main grid and available demand response resources, (3) determines the cost-optimal resource allocation for energy infrastructure, including multiple energy storage systems, and (4) provides a level playing field for emerging technologies, such as power-to-gas and vehicle-to-grid interventions. The multi-energy-storage-technology test-case was effectively applied to achieve 100%-renewable energy generation for the town of Ohakune, New Zealand. Numerical simulation results suggest that the proposed incentive-compatible demand-side management market-clearing mechanism is able to estimate the cost-optimal solution for the provision of renewable energy during the planning phase. The cost-optimal system saves ~21% (equating to around US$5.5 m) compared to a business-as-usual approach, where the participation of end-users in demand response programmes is projected by running uniform price demand response auctions. The most salient distinction of the proposed two-stage (wholesale and retail) demand-side management market model is the continual process of trading, with incentive prices unique to each transaction. [ABSTRACT FROM AUTHOR]