Computed tomography (CT) is widely used in screening, diagnosis, and image-guided therapy for both clinical and research purposes. Since CT involves ionizing radiation, an overarching thrust of related technical research is development of novel methods enabling ultrahigh quality imaging with fine structural details while reducing the X-ray radiation. In this paper, we present a semi-supervised deep learning approach to accurately recover high-resolution (HR) CT images from lowresolution (LR) counterparts. Specifically, with the generative adversarial network (GAN) as the building block, we enforce the cycle-consistency in terms of the Wasserstein distance to establish a nonlinear end-to-end mapping from noisy LR input images to denoised and deblurred HR outputs. We also include the joint constraints in the loss function to facilitate structural preservation. In this deep imaging process, we incorporate deep convolutional neural network (CNN), residual learning, and network in network techniques for feature extraction and restoration. In contrast to the current trend of increasing network depth and complexity to boost the CT imaging performance, which limit its real-world applications by imposing considerable computational and memory overheads, we apply a parallel 1 × 1 1 × 1 1 × 1 CNN to compress the output of the hidden layer and optimize the number of layers and the number of filters for each convolutional layer. Quantitative and qualitative evaluations demonstrate that our proposed model is accurate, efficient and robust for superresolution (SR) image restoration from noisy LR input images. In particular, we validate our composite SR networks on three largescale CT datasets, and obtain promising results as compared to the other state-of-the-art methods.

How to generate testing scenario library for connected and automated vehicles (CAVs) is a major challenge faced by the industry. In previous studies, to evaluate maneuver challenge of a scenario, surrogate models (SMs) are often used without explicit knowledge of the CAV under test. However, performance dissimilarities between the SM and the CAV under test usually exist, and it can lead to the generation of suboptimal scenario library. In this paper, an adaptive testing scenario library generation (ATSLG) method is proposed to solve this problem. A customized testing scenario library for a specific CAV model will be generated as the result of the adaptive process. To estimate the performance dissimilarities and leverage each test of the CAV, Bayesian optimization techniques are applied with classification-based Gaussian Process Regression and a newdesigned acquisition function. Comparing with a pre-determined library, a CAV can be tested and evaluated in a more efficient manner with the customized library. To validate the proposed method, a cut-in and a highway exit case are studied for safety and functionality evaluation respectively. For both two cases, the proposed method can further accelerate the evaluation process by a few orders of magnitudes.

In this paper, we propose a new cooperative driving strategy for connected and automated vehicles (CAVs) at unsignalized intersections. Based on the tree representation of the solution space for the passing order, we combine Monte Carlo tree search (MCTS) and some heuristic rules to find a nearly global-optimal passing order (leaf node) within a very short planning time. Testing results show that this new strategy can keep a good tradeoff between performance and computation flexibility.Index Terms-Connected and Automated Vehicles (CAVs), cooperative driving, unsignalized intersection, Monte Carlo tree search (MCTS).

This paper discusses the trend modeling for traffic time series. First, we recount two types of definitions for a long-term trend that appeared in previous studies and illustrate their intrinsic differences. We show that, by assuming an implicit temporal connection among the time series observed at different days/locations, the PCA trend brings several advantages to traffic time series analysis. We also describe and define the so-called short-term trend that cannot be characterized by existing definitions. Second, we sequentially review the role that trend modeling plays in four major problems in traffic time series analysis: abnormal data detection, data compression, missing data imputation, and traffic prediction. The relations between these problems are revealed, and the benefit of detrending is explained. For the first three problems, we summarize our findings in the last ten years and try to provide an integrated framework for future study. For traffic prediction problem, we present a new explanation on why prediction accuracy can be improved at data points representing the short-term trends if the traffic information from multiple sensors can be appropriately used. This finding indicates that the trend modeling is not only a technique to specify the temporal pattern but is also related to the spatial relation of traffic time series.

Tooth-marked tongue or crenated tongue can provide valuable diagnostic information for traditional Chinese Medicine doctors. However, tooth-marked tongue recognition is challenging. The characteristics of different tongues are multiform and have a great amount of variations, such as different colors, different shapes, and different types of teeth marks. The regions of teeth mark only appear along the lateral borders. Most existing methods make use of concave regions information to classify the tooth-marked tongue which leads to inconstant performance when the region of teeth mark is not concave. In this paper, we try to solve these problems by proposing a three-stage approach which first makes use of concavity information to propose the suspected regions, then use a convolutional neural network to extract deep features and at last use a multiple-instance classifier to make the final decision. Experimental results demonstrate the effectiveness of the proposed method.

No abstract

No abstract

No abstract