This paper deals with dense optical flow estimation from the perspective of the trade-off between quality of the estimated flow and computational cost which is required by real-world applications. We propose a fast and robust local method, denoted by eFOLKI, and describe its implementation on GPU. It leads to very high performance even on large image formats such as 4 K (3,840 × 2,160) resolution. In order to assess the interest of eFOLKI, we first present a comparative study with currently available GPU codes, including local and global methods, on a large set of data with ground truth. eFOLKI appears significantly faster while providing quite accurate and highly robust estimated flows. We then show, on four real-time video processing applications based on optical flow, that eFOLKI reaches the requirements both in terms of estimated flows quality and of processing rate. 相似文献
The vehicle routing problem with deliveries and pickups is one of the main problems within reverse logistics. This paper focuses on an important assumption that divides the literature on the topic, namely the restriction that all deliveries must be completed before pickups can be made. A generalised model is presented, together with a mathematical formulation and its resolution. The latter is carried out by adopting a suitable implementation of the reactive tabu search metaheuristic. Results show that significant savings can be achieved by allowing a mixture of delivery and pickup loads on-board and yet not incurring delays and driver inconvenience. 相似文献
Carboxymethylcellulose (CMC) and beta-cyclodextrin (beta-CD)-based polymers functionalized with two types of quaternary ammonium compounds (QACs), the alkaquat DMB-451 (N-alkyl (50% C14, 40% C12, 10% C10) dimethylbenzylammonium chloride) (DMD-451) named polymer DMB-451, and FMB 1210-8 (a blend of 32 w% N-alkyl (50% C14, 40% C12, 10% C10) dimethylbenzylammonium chloride and 48 w% of didecyldimethylammonium chloride) named polymer FMB 1210-8, were synthethized and characterized by Fourier transform infrared spectroscopy. The antimicrobial activities of these polymers against Eschericia coli were also evaluated at 25 degrees C in wastewater. The results have indicated that the polymer FMB 1210-8 possesses a high-affinity binding with bacterial cells that induces a rapid disinfection process. Moreover, in the same experimental conditions of disinfection (mixture of 1.0 g of polymer and 100 mL of wastewater), the polymer FMB 1210-8 has a higher antimicrobial efficiency (99.90%) than polymer DMB-451 (92.8%). This phenomenon might be associated to a stronger interaction with bacterial cells due to stronger binding affinity for E. coli cells and greater killing efficiency of the C10 alkyl chains QAC of polymer FMB 1210-8 to disrupt the bacterial cell membrane as compared to N-alkyl (50% C14, 40% C12, 10% C10) dimethylbenzylammonium chloride. Together, these results suggest that the polymer FMB 1210-8 could constitute a good disinfectant against Escherichia coli, which could be advantageously used in wastewater treatments due to the low toxicity of beta-CD and CMC, and moderated toxicity of FMB 1210-8 to human and environment. 相似文献
There is no doubt that clustering is one of the most studied data mining tasks. Nevertheless, it remains a challenging problem to solve despite the many proposed clustering approaches. Graph-based approaches solve the clustering task as a global optimization problem, while many other works are based on local methods. In this paper, we propose a novel graph-based algorithm “GBR” that relaxes some well-defined method even as improving the accuracy whilst keeping it simple. The primary motivation of our relaxation of the objective is to allow the reformulated objective to find well distributed cluster indicators for complicated data instances. This relaxation results in an analytical solution that avoids the approximated iterative methods that have been adopted in many other graph-based approaches. The experiments on synthetic and real data sets show that our relaxation accomplishes excellent clustering results. Our key contributions are: (1) we provide an analytical solution to solve the global clustering task as opposed to approximated iterative approaches; (2) a very simple implementation using existing optimization packages; (3) an algorithm with relatively less computation time over the number of data instances to cluster than other well defined methods in the literature. 相似文献
Hierarchical clustering is a stepwise clustering method usually based on proximity measures between objects or sets of objects from a given data set. The most common proximity measures are distance measures. The derived proximity matrices can be used to build graphs, which provide the basic structure for some clustering methods. We present here a new proximity matrix based on an entropic measure and also a clustering algorithm (LEGCIust) that builds layers of subgraphs based on this matrix and uses them and a hierarchical agglomerative clustering technique to form the clusters. Our approach capitalizes on both a graph structure and a hierarchical construction. Moreover, by using entropy as a proximity measure, we are able, with no assumption about the cluster shapes, to capture the local structure of the data, forcing the clustering method to reflect this structure. We present several experiments on artificial and real data sets that provide evidence on the superior performance of this new algorithm when compared with competing ones. 相似文献
Hierarchical clustering is a stepwise clustering method usually based on proximity measures between objects or sets of objects from a given data set. The most common proximity measures are distance measures. The derived proximity matrices can be used to build graphs, which provide the basic structure for some clustering methods. We present here a new proximity matrix based on an entropic measure and also a clustering algorithm (LEGClust) that builds layers of subgraphs based on this matrix, and uses them and a hierarchical agglomerative clustering technique to form the clusters. Our approach capitalizes on both a graph structure and a hierarchical construction. Moreover, by using entropy as a proximity measure we are able, with no assumption about the cluster shapes, to capture the local structure of the data, forcing the clustering method to reflect this structure. We present several experiments on artificial and real data sets that provide evidence on the superior performance of this new algorithm when compared with competing ones. 相似文献
Real-time crowd motion planning requires fast, realistic methods for path planning as well as obstacle avoidance. In a previous
work (Morini et al. in Cyberworlds International Conference, pp. 144–151, 2007), we introduced a hybrid architecture to handle real-time motion planning of thousands of pedestrians. In this article, we
present an extended version of our architecture, introducing two new features: an improved short-term collision avoidance
algorithm, and simple efficient group behavior for crowds. Our approach allows the use of several motion planning algorithms
of different precision for regions of varied interest. Pedestrian motion continuity is ensured when switching between such
algorithms. To assess our architecture, several performance tests have been conducted, as well as a subjective test demonstrating
the impact of using groups. Our results show that the architecture can plan motion in real time for several thousands of characters.
The joint estimation of the location vector and the shape matrix of a set of independent and identically Complex Elliptically Symmetric (CES) distributed observations is investigated from both the theoretical and computational viewpoints. This joint estimation problem is framed in the original context of semiparametric models allowing us to handle the (generally unknown) density generator as an infinite-dimensional nuisance parameter. In the first part of the paper, a computationally efficient and memory saving implementation of the robust and semiparmaetric efficient R-estimator for shape matrices is derived. Building upon this result, in the second part, a joint estimator, relying on the Tyler’s M-estimator of location and on the R-estimator of shape matrix, is proposed and its Mean Squared Error (MSE) performance compared with the Semiparametric Cramér-Rao Bound (SCRB).