Search

Advanced Driving Assistance System (ADAS) can predict pedestrian's trajectory, in order to avoid traffic accidents and guarantee driving safety. A few current pedestrian trajectory prediction methods use a pedestrian's historical motion to predict the future trajectory, but the pedestrian's trajectory is also affected by the vehicle using the ADAS for prediction (target vehicle). Other studies predict the pedestrian's and vehicle's trajectories separately, and use the latter to adjust the former, but their interaction is a continuous process and should be considered during prediction rather than after. Therefore, we propose PVII, a pedestrian-vehicle interactive and iterative prediction framework for pedestrian's trajectory. It makes prediction for one iteration based on the results from previous iteration, which essentially models the vehicle-pedestrian interaction. In this iterative framework, to avoid accumulation of prediction errors along with the increased iterations, we design a bi-layer Bayesian en/decoder. For each iteration, it not only uses inaccurate results from previous iteration but also accurate historical data for prediction, and calculates Bayesian uncertainty values to evaluate the results. In addition, the pedestrian's trajectory is affected by both target vehicle and other vehicles around it (surrounding vehicle), so we include into the framework a pre-trained speed estimation module for surrounding vehicles (SE module). It estimates the speed based on pedestrian's motion and we collect data from pedestrian's view for training. In experiments, PVII can achieve the highest prediction accuracy compared to the current methods. [ABSTRACT FROM AUTHOR]

This paper presents two novel solution approaches for addressing large-scale multi-floor process plant layout problems. Based on the mixed integer linear programming (MILP) model, the first solution approach employs a multi-directional search strategy while the second improves solution efficiency by reducing model size through an iterative framework. Both approaches determine the spatial arrangement of the plant equipment considering equipment-floor allocation, non-overlapping constraints, tall equipment penetrating multiple floors, etc. The computational results indicate that the proposed approaches achieved potential cost savings for four illustrative examples when compared to the previous studies. Finally, engineering experience constraints were included to represent a more complex industrial situation, and their applicability was tested with the last example. [ABSTRACT FROM AUTHOR]

Influence maximization for opinion formation (IMOF) in social networks is an important problem, which is used to determine some initial nodes and propagate the most ideal opinions to the whole network. The existing researches focus on improving the opinion formation models to compute the opinion of each node. However, little work has been done to describe the IMOF process mathematically, and the current researches cannot provide an effective mechanism to deal with the IMOF. In this paper, the IMOF is formulated mathematically and solved by an iterative framework. At first, we describe the IMOF as a constrained optimization problem. Then, based on node influence and neighbor coordination, the weighted coordination model is proposed to compute the opinions of network nodes with the change of iterations. In particular, in order to determine top- $k$ influential nodes (i.e., seed nodes), an iterative framework for the IMOF, called IIMOF is presented. Based on the framework, the score and rank of each node by Iterative 2-hop algorithm, i.e., SRI2 is proposed to compute the influence score of each node. Based on small in-degree and high out-degree, one-hop measure is proposed to better reflect the rank of all initial nodes. We also prove that IIMOF converges to a stable order set within the finite iterations. The simulation results show that IIMOF has superior average opinions than the comparison algorithms.

Highlights • We propose a two-stage iterative framework for influence maximization in social networks (ie., TIFIM). It combines the spread benefit with selection of seed nodes, guaranteeing the remarkable efficiency as well as high accuracy. • Based on the last iteration results and the two-hop measure, we put forward an efficient FLAS to calculate spread benefit of each node, further improving the efficiency and accuracy of TIFIM. • We define the apical dominance to describe the overlapping phenomenon among nodes. We further propose RAD to determine the seed nodes from candidate nodes. Abstract Influence Maximization is an important problem in social networks, and its main goal is to select some most influential initial nodes (i.e., seed nodes) to obtain the maximal influence spread. The existing studies primarily concentrate on the corresponding methods for influence maximization, including greedy algorithms, heuristic algorithms and their extensions to determine the most influential nodes. However, there is little work to ensure efficiency and accuracy of the proposed schemes at the same time. In this paper, a Two-stage Iterative Framework for the Influence Maximization in social networks, (i.e., TIFIM) is proposed. In order to exclude less influential nodes and decrease the computation complexity of TIFIM, in the first stage, an iterative framework in descending order is proposed to select the candidate nodes. In particular, based on the results of the last iteration and the two-hop measure, the First-Last Allocating Strategy (FLAS) is presented to compute the spread benefit of each node. We prove that TIFIM converges to a stable order within the finite iterations. In the second stage, we define the apical dominance to calculate the overlapping phenomenon of spread benefit among nodes and further propose Removal of the Apical Dominance (RAD) to determine seed nodes from the candidate nodes. Moreover, we also prove that the influence spread of TIFIM according to RAD converges to a specific value within finite computations. Finally, simulation results show that the proposed scheme has superior influence spread and running time than other existing ones. [ABSTRACT FROM AUTHOR]

The interconnection and synchronization among different transport modes have been more and more attractive as the modern transportation system is moving towards Mobility-as-a-Service. In this study, we address the train scheduling problem for "commuter rail-metro" systems, where the trains from commuter rail lines can go directly into metro systems to provide seamless services for passengers. To optimize the schedule of trains for both commuter rail lines and metro lines, we propose a job shop scheduling model where precedence constraints from commuter-metro networks are taken into account and develop a mixed-integer programming (MIP) model with quadratic constraints. Our model considers the orders of different types of trains and the safety constraints, due to different types of signalling equipment in commuter and metro systems. Since these constraints involve a set of IF-THEN rules, we prove that these constraints can be equivalently reformulated as linear inequalities, without adding new variables. To solve the proposed model efficiently, we design an iterative solution framework, which generates a feasible solution using dynamic programming, next solves a MIP model, then calculates the train speed profiles, and if train speed profiles violate the safety constraints, re-optimizes the MIP model with modified alternative constraints. To verify the effectiveness of the proposed approaches, numerical experiments are performed on small and real-world instances based on the Beijing metro Line 1 and the Batong Line operational data. • Train scheduling and train speed management in commuter-metro networks. • A mixed-integer quadratic programming model based on the job shop scheduling. • We propose an iterative solution methodology for our integrated scheduling problem. • Results verify the effectiveness of our approaches and derive managerial insights. [ABSTRACT FROM AUTHOR]

Recent works have proposed two L1-norm distance measure-based linear discriminant analysis (LDA) methods, L1-LD and LDA-L1, which aim to promote the robustness of the conventional LDA against outliers. In LDA-L1, a gradient ascending iterative algorithm is applied, which, however, suffers from the choice of stepwise. In L1-LDA, an alternating optimization strategy is proposed to overcome this problem. In this paper, however, we show that due to the use of this strategy, L1-LDA is accompanied with some serious problems that hinder the derivation of the optimal discrimination for data. Then, we propose an effective iterative framework to solve a general L1-norm minimization–maximization (minmax) problem. Based on the framework, we further develop a effective L1-norm distance-based LDA (called L1-ELDA) method. Theoretical insights into the convergence and effectiveness of our algorithm are provided and further verified by extensive experimental results on image databases. [ABSTRACT FROM PUBLISHER]

Automatic to locate the salient regions in the images are useful for many computer vision and computer graphics tasks. However, the previous techniques prefer to give noisy and fuzzy saliency maps, which will be a crucial limitation for the performance of subsequent image processing. In this paper, we present a novel framework by aggregating various bottom-up cues and bias to enhance visual saliency detection. It can produce high-resolution, full-field saliency map which can be close to binary one and more effective in real-world applications. First, the proposed method concentrates on multiple saliency cues in a global context, such as regional contrast, spatial relationship and color histogram smoothing to produce a coarse saliency map. Second, combining complementary boundary prior with smoothing, we iteratively refine the coarse saliency map to improve the contrast between salient and non-salient regions until a close to binary saliency map is reached. Finally, we evaluate our salient region detection on two publicly available datasets with pixel accurate annotations. The experimental results show that the proposed method performs equally or better than the 12 alternative methods and retains comparable detection accuracy, even in extreme cases. Furthermore, we demonstrate that the saliency map produced by our approach can serve as a good initialization for automatic alpha matting and image retargeting. [ABSTRACT FROM AUTHOR]

Spatial co-location pattern mining discovers the subsets of features whose events are frequently located together in geographic space. The current research on this topic adopts a threshold-based approach that requires users to specify in advance the thresholds of distance and prevalence. However, in practice, it is not easy to specify suitable thresholds. In this article, we propose a novel iterative mining framework that discovers spatial co-location patterns without predefined thresholds. With the absolute and relative prevalence of spatial co-locations, our method allows users to iteratively select informative edges to construct the neighborhood relationship graph until every significant co-location has enough confidence and eventually to discover all spatial co-location patterns. The experimental results on real world data sets indicate that our framework is effective for prevalent co-locations discovery. [ABSTRACT FROM AUTHOR]