We tackle the problem of large-scale object detection in images, where the number of objects can be arbitrarily large, and can exhibit significant overlap/occlusion. A successful approach to modelling the large-scale ...
详细信息
ISBN:
(纸本)9781467388528
We tackle the problem of large-scale object detection in images, where the number of objects can be arbitrarily large, and can exhibit significant overlap/occlusion. A successful approach to modelling the large-scale nature of this problem has been via point process density functions which jointly encode object qualities and spatial interactions. But the corresponding optimisation problem is typically difficult or intractable, and many of the best current methods rely on Monte Carlo Markov Chain (MCMC) simulation, which converges slowly in a large solution space. We propose an efficient point process inference for large-scale object detection using discrete energy minimization. In particular, we approximate the solution space by a finite set of object proposals and cast the point process density function to a corresponding energy function of binary variables whose values indicate which object proposals are accepted. We resort to the local submodular approximation (LSA) based trust-region optimisation to find the optimal solution. Furthermore we analyse the error of LSA approximation, and show how to adjust the point process energy to dramatically speed up the convergence without harming the optimality. We demonstrate the superior efficiency and accuracy of our method using a variety of large-scale object detection applications such as crowd human detection, birds, cells counting/localization.
This paper presents an optimized solution of finding time-optimal trajectories for autonomous systems. These systems are subject to avoidance requirements, which include avoidance of collisions with other systems and ...
详细信息
ISBN:
(纸本)9781424418060
This paper presents an optimized solution of finding time-optimal trajectories for autonomous systems. These systems are subject to avoidance requirements, which include avoidance of collisions with other systems and obstacles, either static or dynamic. The necessary constraints for avoidance are added to a time-optimizing linear program by including a binary variable in the optimization. The resulting problem is a mixed-integer linear program (MILP). This will be solved using AMPL mathematical programming language in conjunction with CPLEX optimization software.
The current paper addresses the problem of optimizing a cost function over a non-convex and possibly non-connected feasible region. A classical approach for solving this type of optimization problem is based on Mixed ...
详细信息
ISBN:
(纸本)9781612848006
The current paper addresses the problem of optimizing a cost function over a non-convex and possibly non-connected feasible region. A classical approach for solving this type of optimization problem is based on Mixed integer technique. The exponential complexity as a function of the number of binary variables used in the problem formulation highlights the importance of reducing them. Previous work which minimize the number of binary variables is revisited and enhanced. Practical limitations of the procedure are discussed and a typical control application, the control of Multi-Agent Systems is exemplified.
The bounding box representation employed by many popular object detection models [3, 6] implicitly assumes all pixels inside the box belong to the object. This assumption makes this representation less robust to the o...
详细信息
ISBN:
(纸本)9781457703942
The bounding box representation employed by many popular object detection models [3, 6] implicitly assumes all pixels inside the box belong to the object. This assumption makes this representation less robust to the object with occlusion [16]. In this paper, we augment the bounding box with a set of binary variables each of which corresponds to a cell indicating whether the pixels in the cell belong to the object. This segmentation-aware representation explicitly models and accounts for the supporting pixels for the object within the bounding box thus more robust to occlusion. We learn the model in a structured output framework, and develop a method that efficiently performs both inference and learning using this rich representation. The method is able to use segmentation reasoning to achieve improved detection results with richer output (cell level segmentation) on the Street Scenes and Pascal VOC 2007 datasets. Finally, we present a globally coherent object model using our rich representation to account for object-object occlusion resulting in a more coherent image understanding.
Access to information is key for improving the position of persons with disabilities in society. Familiarity with state regulations regarding access to information could be influenced by communication with state autho...
详细信息
Access to information is key for improving the position of persons with disabilities in society. Familiarity with state regulations regarding access to information could be influenced by communication with state authorities concerning the rights of persons with disabilities, especially access to information. Familiarity with these regulations and the specified communication with state authorities might be affected by a number of background variables, such as age and education completed. To clarify relations among these variables, which would enable state authorities and other relevant institutions to define and implement policies that might improve matters, there is a need to prepare and analyze appropriate datasets concerning them. This paper describes such a dataset, preliminary in nature, obtained from answers to part of a questionnaire administered to persons with disabilities living in Serbia. Persons with innate or acquired physical and/or sensory disability were included in the research. This dataset contains raw data of nine variables, as well as analyzed data of ten variables derived from most of the raw data. Besides correlative analyses, the dataset was previously analyzed using PLS (partial least squares) path modeling. To reuse the dataset, a path model with Bayesian estimations may be applied, whose outcomes for different model priors (prior distributions) may be compared to those of the PLS path modeling. The dataset also contains data of two variables that may be included in further research. (C) 2020 The Authors. Published by Elsevier Inc.
Despite known shortcomings of the procedure, exploratory factor analysis of dichotomous test items has been limited, until recently, to unweighted analyses of matrices of tetrachoric correlations. Superior methods hav...
详细信息
Despite known shortcomings of the procedure, exploratory factor analysis of dichotomous test items has been limited, until recently, to unweighted analyses of matrices of tetrachoric correlations. Superior methods have begun to appear in the literature, in professional symposia, and in computer programs. This paper places these developments in a unified framework, from a review of the classical common factor model for measured variables through generalized least squares and marginal maximum likelihod solutions for dichotomous data. Further extensions of the model are also reported as work in progress.
暂无评论