摘要 :
Optimal transport is a machine learning problem with applications including distribution comparison, feature selection, and generative adversarial networks. In this paper, we propose feature-robust optimal transport (FROT) for hig...
展开
Optimal transport is a machine learning problem with applications including distribution comparison, feature selection, and generative adversarial networks. In this paper, we propose feature-robust optimal transport (FROT) for high-dimensional data, which solves high-dimensional OT problems using feature selection to avoid the curse of dimensionality. Specifically, we find a transport plan with discriminative features. To this end, we formulate the FROT problem as a min-max optimization problem. We then propose a convex formulation of the FROT problem and solve it using a Frank-Wolfe-based optimization algorithm, whereby the subproblem can be efficiently solved using the Sinkhorn algorithm. Since FROT finds the transport plan from selected features, it is robust to noise features. To show the effectiveness of FROT, we propose using the FROT algorithm for the layer selection problem in deep neural networks for semantic correspondence. By conducting synthetic and benchmark experiments, we demonstrate that the proposed method can find a strong correspondence by determining important layers. We show that the FROT algorithm achieves state-of-the-art performance in real-world semantic correspondence datasets. Code can be found at https://github.com/Mathux/FROT.
收起
摘要 :
Estimating mutual information is an important statistics and machine learning problem. To estimate the mutual information from data, a common practice is preparing a set of paired samples {(x_i, y_i)}~n_i=1 ～ p(x, y). However, in...
展开
Estimating mutual information is an important statistics and machine learning problem. To estimate the mutual information from data, a common practice is preparing a set of paired samples {(x_i, y_i)}~n_i=1 ～ p(x, y). However, in many situations, it is difficult to obtain a large number of data pairs. To address this problem, we propose the semi-supervised Squared-loss Mutual Information (SMI) estimation method using a small number of paired samples and the available unpaired ones. We first represent SMI through the density ratio function, where the expectation is approximated by the samples from marginals and its assignment parameters. The objective is formulated using the optimal transport problem and quadratic programming. Then, we introduce the Least-Squares Mutual Information with Sinkhorn (LSMI-Sinkhorn) algorithm for efficient optimization. Through experiments, we first demonstrate that the proposed method can estimate the SMI without a large number of paired samples. Then, we show the effectiveness of the proposed LSMI-Sinkhorn algorithm on various types of machine learning problems such as image matching and photo album summarization. Code can be found at https://github.com/csyanbin/LSMI-Sinkhorn.
收起
摘要 :
Estimating mutual information is an important statistics and machine learning problem. To estimate the mutual information from data, a common practice is preparing a set of paired samples {(x_i,y_i)}_(i=1)~n ~(i.i.d.)～ p(x, y). H...
展开
Estimating mutual information is an important statistics and machine learning problem. To estimate the mutual information from data, a common practice is preparing a set of paired samples {(x_i,y_i)}_(i=1)~n ~(i.i.d.)～ p(x, y). However, in many situations, it is difficult to obtain a large number of data pairs. To address this problem, we propose the semi-supervised Squared-loss Mutual Information (SMI) estimation method using a small number of paired samples and the available unpaired ones. We first represent SMI through the density ratio function, where the expectation is approximated by the samples from marginals and its assignment parameters. The objective is formulated using the optimal transport problem and quadratic programming. Then, we introduce the Least-Squares Mutual Information with Sinkhorn (LSMI-Sinkhorn) algorithm for efficient optimization. Through experiments, we first demonstrate that the proposed method can estimate the SMI without a large number of paired samples. Then, we show the effectiveness of the proposed LSMI-Sinkhorn algorithm on various types of machine learning problems such as image matching and photo album summarization.
收起
摘要 :
We present a general model of interdomain route selection to study interdomain traffic engineering. In this model, the routing of multiple destinations can be coordinated. Thus the model can capture general traffic engineering beh...
展开
We present a general model of interdomain route selection to study interdomain traffic engineering. In this model, the routing of multiple destinations can be coordinated. Thus the model can capture general traffic engineering behaviors such as load balancing and link capacity constraints. We first identify potential routing instability and inefficiency of interdomain traffic engineering. We then derive a sufficient condition to guarantee convergence. We also show that the constraints on local policies imposed by business considerations in the Internet can guarantee stability without global coordination. Using realistic Internet topology, we evaluate the extent to which routing instability of interdomain traffic engineering can happen when the constraints are violated.
收起
摘要 :
Traffic accident forecasting is of vital importance to the intelligent transportation and public safety. Spatial-temporal learning is the mainstream approach to exploring complex evolving patterns. However, two intrinsic challenge...
展开
Traffic accident forecasting is of vital importance to the intelligent transportation and public safety. Spatial-temporal learning is the mainstream approach to exploring complex evolving patterns. However, two intrinsic challenges lie in traffic accident forecasting, preventing the straightforward adoption of spatial-temporal learning. First, the temporal observations of traffic accidents exhibit ultra-rareness due to the inherent properties of accident occurrences (Fig. 1(a)), which leads to the severe scarcity of risk samples in learning accident patterns. Second, the spatial distribution of accidents is severely imbalanced from region to region (Fig. 1(b)), which poses a serious challenge to forecast the spatially diversified risks. To tackle the above challenges, we propose RiskContra, a Contrastive learning approach with multi-kernel networks, to forecast the Risk of traffic accidents. Specifically, to address the first challenge (i.e. temporal rareness), we design a novel contrastive learning approach, which leverages the periodic patterns to derive a tailored mixup strategy for risk sample augmentation. This way, the contrastively learned features can better represent the risk samples, thus capturing higher-quality accident patterns for forecasting. To address the second challenge (i.e. spatial imbalance), we design the multi-kernel networks to capture the hierarchical correlations from multiple spatial granularities. This way, disparate regions can utilize the multi-granularity correlations to enhance the forecasting performance across regions. Extensive experiments corroborate the effectiveness of each devised component in RiskContra.
收起
摘要 :
The model of thread connection preload is set up and the main factors affecting the performance of thread axial preload are analyzed theoretically. Based on the steel-aluminum connection structure, the test was carried out on the ...
展开
The model of thread connection preload is set up and the main factors affecting the performance of thread axial preload are analyzed theoretically. Based on the steel-aluminum connection structure, the test was carried out on the meshing length of aluminum thread, the friction of aluminum support surface, the friction of aluminum thread and the rotation speed by using the tightening system and ultrasonic preload measuring instrument. The results show that the stability of axial preload increases with the increase of the meshing length of aluminum thread. Increase the plain washer to effectively improve the friction state of the supporting surface, axial preload increased by nearly 40%; The friction state of thread was effectively improved by adding wire thread sleeve, and the axial preload was increased by nearly 46%. It is suggested that the proper increase of rotation speed is of certain significance to ensure the reliability of threaded connections.
收起
摘要 :
In this paper, we proposed an adaptive enhancement algorithm based on fuzzy relaxation. Firstly, the OTSU algorithm is used to classify background and objective, and the crossover points for each pixel are defined by the classific...
展开
In this paper, we proposed an adaptive enhancement algorithm based on fuzzy relaxation. Firstly, the OTSU algorithm is used to classify background and objective, and the crossover points for each pixel are defined by the classification results. Then, the concept of fuzzy contrast based on the image normalization is introduced, and the value of fuzzy contrast is defined as a image contrast feature plane. Secondly, at basis of the fuzzy characteristic of the hyperbolic tangent, a novel membership function is proposed, the crossover points and the adaptive function curve can achieve the best by adjusting the control parameters. Finally, the fuzzy contrast feature plane is mapped to gray level plane using the method of linear transformation. The experiment obtains excellent results which is only one time iteration. The linear transformation reduces the lose of the adjacent materials bag image's edge information and improves the operational efficiency. The analysis experimentally demonstrates that proposed algorithm is adaptive and the image details also have been preserved.
收起
摘要 :
In this paper, we proposed an adaptive enhancement algorithm based on fuzzy relaxation. Firstly, the OTSU algorithm is used to classify background and objective, and the crossover points for each pixel are defined by the classific...
展开
In this paper, we proposed an adaptive enhancement algorithm based on fuzzy relaxation. Firstly, the OTSU algorithm is used to classify background and objective, and the crossover points for each pixel are defined by the classification results. Then, the concept of fuzzy contrast based on the image normalization is introduced, and the value of fuzzy contrast is defined as a image contrast feature plane. Secondly, at basis of the fuzzy characteristic of the hyperbolic tangent, a novel membership function is proposed, the crossover points and the adaptive function curve can achieve the best by adjusting the control parameters. Finally, the fuzzy contrast feature plane is mapped to gray level plane using the method of linear transformation. The experiment obtains excellent results which is only one time iteration. The linear transformation reduces the lose of the adjacent materials bag image’s edge information and improves the operational efficiency. The analysis experimentally demonstrates that proposed algorithm is adaptive and the image details also have been preserved.
收起
摘要 :
In this paper, we proposed an adaptive enhancement algorithm based on fuzzy relaxation. Firstly, the OTSU algorithm is used to classify background and objective, and the crossover points for each pixel are defined by the classific...
展开
In this paper, we proposed an adaptive enhancement algorithm based on fuzzy relaxation. Firstly, the OTSU algorithm is used to classify background and objective, and the crossover points for each pixel are defined by the classification results. Then, the concept of fuzzy contrast based on the image normalization is introduced, and the value of fuzzy contrast is defined as a image contrast feature plane. Secondly, at basis of the fuzzy characteristic of the hyperbolic tangent, a novel membership function is proposed, the crossover points and the adaptive function curve can achieve the best by adjusting the control parameters. Finally, the fuzzy contrast feature plane is mapped to gray level plane using the method of linear transformation. The experiment obtains excellent results which is only one time iteration. The linear transformation reduces the lose of the adjacent materials bag image's edge information and improves the operational efficiency. The analysis experimentally demonstrates that proposed algorithm is adaptive and the image details also have been preserved.
收起
摘要 :
In this paper, we proposed an adaptive enhancement algorithm based on fuzzy relaxation. Firstly, the OTSU algorithm is used to classify background and objective, and the crossover points for each pixel are defined by the classific...
展开
In this paper, we proposed an adaptive enhancement algorithm based on fuzzy relaxation. Firstly, the OTSU algorithm is used to classify background and objective, and the crossover points for each pixel are defined by the classification results. Then, the concept of fuzzy contrast based on the image normalization is introduced, and the value of fuzzy contrast is defined as a image contrast feature plane. Secondly, at basis of the fuzzy characteristic of the hyperbolic tangent, a novel membership function is proposed, the crossover points and the adaptive function curve can achieve the best by adjusting the control parameters. Finally, the fuzzy contrast feature plane is mapped to gray level plane using the method of linear transformation. The experiment obtains excellent results which is only one time iteration. The linear transformation reduces the lose of the adjacent materials bag image's edge information and improves the operational efficiency. The analysis experimentally demonstrates that proposed algorithm is adaptive and the image details also have been preserved.
收起
原文获取