Search

Abstract A more complete taxi pick‐ups prediction going beyond the conditional expectation would be more beneficial to allocate taxis effectively. Prior works have predicted the conditional expectation of taxi demand and improved the prediction accuracy. But the ability to convey calibrated uncertainty estimates in prediction was less studied. Here, a multi‐output multi‐quantile deep neural network architecture is presented for predicting the mean and the multiple quantiles simultaneously, which jointly models multiple quantiles together with the conditional expectation based on self‐attention ConvLSTM. Using the NYC taxi dataset, we have analyzed the performance of the developed architecture. For the mean prediction, the proposed method has an error of 8.584 in terms of RMSE criteria and Linear QR, Mcdropout‐Gal, ConvLSTM and SA‐ConvLSTM methods, which have error values of 8.800, 8.671, 8.599, and 8.597, respectively. According to the MAE criterion, this method has a value of 5.934. The other methods have obtained values of 6.614, 5.990, 5.986, and 5.959, respectively. For multiple quantiles prediction, the proposed method produces narrower intervals while obtaining the target coverage percentage. Overall, the proposed model not only outperforms the other models, but also provides a rich description about the predictive density.

Abstract In this paper, a novel variational model with box constraints is proposed to restore images corrupted by impulse noise. The proposed model is composed of fractional‐order total variation regularization and Lp‐fidelity term (0

It is very common that taxi is used as a feeder mode to/from a metro station especially in third-tier cities of China. But research on taxi-metro integrated use is rather limited. This paper investigates the relationship between feeder-related built environment and taxi-metro integrated use by using taxi trajectory data in Lanzhou, China. Firstly, regression models are developed to explore the transfer distance of access/egress trips during peak hours. Then, the catchment area is delineated for collecting feeder-related environment variables. Finally, several negative binomial regressions are employed to examine the relationships between feeder-related built environment and taxi-metro integrated use during peak hours on weekdays. The results reveal that (a) people prefer to use taxi as feeder mode for metro access during morning peak hours and for metro egress during evening peak hours; (b) the transfer distance of taxi-metro integrated use is about 3.8 km; (c) higher mixed land use generates more taxi-metro integrated use during evening peak hours. Higher proportion of residential land use attracts more taxi-metro integrated use for metro access during morning peak hours. Those findings will help transport planners to develop tailored land-use interventions to improve transit accessibility and promote the sustainable multimodal travel.

It is widely known that the total variation regularization model preserves the edges well in the restored images but has some staircase effects. We consider using non-convex high-order total variation and overlapping group sparsity as a hybrid regularization to present a new denoising model. The proposed model can well preserve edges and reduce the staircase effect in the smooth region of the restored images. In order to solve the proposed hybrid model, we develop an efficient alternating minimization method. Compared with other models for removing Cauchy noise, numerical experimental results demonstrate that the superiority of the proposed model and algorithm, both in terms of visual and quantitative measures.

Multiplicative noise removal is a quite challenging problem in image denoising. In recent years, hyper-Laplacian prior information has been successfully introduced in the image denoising problem and significant denoising effects have been achieved. In this paper, we propose a new hybrid regularizer model for removing multiplicative noise. The proposed model consists of the non-convex higher-order total variation and overlapping group sparsity on a hyper-Laplacian prior regularizer. It combines the advantages of the non-convex regularization and the hybrid regularization, which may simultaneously preserve the fine-edge information and reduce the staircase effect at the same time. We develop an effective alternating minimization method for the proposed nonconvex model via an alternating direction method of multipliers framework, where the majorization–minimization algorithm and the iteratively reweighted algorithm are adopted to solve the corresponding subproblems. Numerical experiments show that the proposed model outperforms the most advanced model in terms of visual quality and certain image quality measurements.

It is widely known that the classic total variation(TV) model has been proven to be very effective in preserving sharp edges. However, the TV model suffers from the staircase effects which produce blocking artifacts in the restored images. In this paper, we propose a new hybrid regularization model by combining non-convex second order total variation with wavelet transform to restrain staircase effects and protect some details of the images. To compute the new model effectively, we propose an alternating minimization method for recovering images from the blurry and noisy observations. The new model is first transformed into several sub-problems, and the generalized iterated shrinkage algorithm, the Fourier transform method and projection method are used to solve these sub-problems, respectively. Numerical experiments show that the proposed model can restrain blocking artifacts while projecting sharp edges, and the restoration quality outperforms several state-of-the-art methods.

In this paper, we propose a new high-order total variation regularized model with box constraint for image compressive sensing reconstruction. Because of the separable structure of this model, we can easily decompose into three subproblems by splitting the augmented Lagrangian function. To effectively solve the proposed new model, a fast alternating minimization method with accelerated technique is presented. Moreover, the proposed method applies a linearized strategy for quadratic terms to get the closed-form solution and reduce the computation cost. Numerical experiments show that our proposed model can get better performance than several current state-of-the-art methods in terms of signal to noise ratio (SNR) and visual perception.

Abstract The total variation model is widely used in image deblurring and denoising process with the features of protecting the image edge. However, this model usually causes some staircase effects. To overcome the shortcoming, combining the second-order total variation regularization and the total variation regularization, we propose a hybrid total variation model. The new improved model not only eliminates the staircase effect, but also well protects the edges of the image. The alternating direction method of multipliers (ADMM) is employed to solve the proposed model. Numerical results show that our proposed model can get more details and higher image visual quality than some current state-of-the-art methods.

In recent years, the fractional-order derivative has achieved great success in removing Gaussian noise, impulsive noise, multiplicative noise and so on, but few works have been conducted to remove Cauchy noise. In this paper, we propose a novel nonconvex variational model for removing Cauchy noise based on the truncated fractional-order total variation. The new model can effectively reduce the staircase effect and keep small details or textures while removing Cauchy noise. In order to solve the nonconvex truncated fractional-order total variation regularization model, we propose an efficient alternating minimization method under the framework of the alternating direction multiplier method. Experimental results illustrate the effectiveness of the proposed model, compared to some previous models.

It has been proved that total generalized variation (TGV) can better preserve edges while suppressing staircase effect. In this paper, we propose an effective hybrid regularization model based on second-order TGV and wavelet frame. The proposed model inherits the advantages of TGV regularization and wavelet frame regularization, can eliminate staircase effect while protecting the sharp edge, and simultaneously has good capability of sparsely estimating the piecewise smooth functions. The alternative direction method of multiplier (ADMM) is employed to solve the new model. Numerical results show that our proposed model can preserve more details and get higher image visual quality than some current state-of-the-art methods.

With the advances in micro-electronics, wireless sensor devices have been made much smaller and more integrated, and large-scale wireless sensor networks (WSNs) based the cooperation among the significant amount of nodes have become a hot topic. “Large-scale” means mainly large area or high density of a network. Accordingly the routing protocols must scale well to the network scope extension and node density increases. A sensor node is normally energy-limited and cannot be recharged, and thus its energy consumption has a quite significant effect on the scalability of the protocol. To the best of our knowledge, currently the mainstream methods to solve the energy problem in large-scale WSNs are the hierarchical routing protocols. In a hierarchical routing protocol, all the nodes are divided into several groups with different assignment levels. The nodes within the high level are responsible for data aggregation and management work, and the low level nodes for sensing their surroundings and collecting information. The hierarchical routing protocols are proved to be more energy-efficient than flat ones in which all the nodes play the same role, especially in terms of the data aggregation and the flooding of the control packets. With focus on the hierarchical structure, in this paper we provide an insight into routing protocols designed specifically for large-scale WSNs. According to the different objectives, the protocols are generally classified based on different criteria such as control overhead reduction, energy consumption mitigation and energy balance. In order to gain a comprehensive understanding of each protocol, we highlight their innovative ideas, describe the underlying principles in detail and analyze their advantages and disadvantages. Moreover a comparison of each routing protocol is conducted to demonstrate the differences between the protocols in terms of message complexity, memory requirements, localization, data aggregation, clustering manner and other metrics. Finally some open issues in routing protocol design in large-scale wireless sensor networks and conclusions are proposed.

By using slack variables and minimum function, we first reformulate the system of equalities and inequalities as a system of nonsmooth equations, and, using smoothing technique, we construct the smooth operator. A new noninterior continuation method is proposed to solve the system of smooth equations. It shows that any accumulation point of the iteration sequence generated by our algorithm is a solution of the system of equalities and inequalities. Some numerical experiments show the feasibility and efficiency of the algorithm.

We propose a new edge preserving image deconvolution model by combining total variation and nonlocal means regularization. Natural images exhibit an high degree of redundancy. Using this redundancy, the nonlocal means regularization strategy is a good technique for detail preserving image restoration. In order to further improve the visual quality of the nonlocal means based algorithm, total variation is introduced to the model to better preserve edges. Then an efficient alternating minimization procedure is used to solve the model. Numerical experiments illustrate the effectiveness of the proposed algorithm.

With the advances in micro-electronics, wireless sensor devices have been made much smaller and more integrated, and large-scale wireless sensor networks (WSNs) based the cooperation among the significant amount of nodes have become a hot topic. "Large-scale" means mainly large area or high density of a network. Accordingly the routing protocols must scale well to the network scope extension and node density increases. A sensor node is normally energy-limited and cannot be recharged, and thus its energy consumption has a quite significant effect on the scalability of the protocol. To the best of our knowledge, currently the mainstream methods to solve the energy problem in large-scale WSNs are the hierarchical routing protocols. In a hierarchical routing protocol, all the nodes are divided into several groups with different assignment levels. The nodes within the high level are responsible for data aggregation and management work, and the low level nodes for sensing their surroundings and collecting information. The hierarchical routing protocols are proved to be more energy-efficient than flat ones in which all the nodes play the same role, especially in terms of the data aggregation and the flooding of the control packets. With focus on the hierarchical structure, in this paper we provide an insight into routing protocols designed specifically for large-scale WSNs. According to the different objectives, the protocols are generally classified based on different criteria such as control overhead reduction, energy consumption mitigation and energy balance. In order to gain a comprehensive understanding of each protocol, we highlight their innovative ideas, describe the underlying principles in detail and analyze their advantages and disadvantages. Moreover a comparison of each routing protocol is conducted to demonstrate the differences between the protocols in terms of message complexity, memory requirements, localization, data aggregation, clustering manner and other metrics. Finally some open issues in routing protocol design in large-scale wireless sensor networks and conclusions are proposed. [ABSTRACT FROM AUTHOR]