Search

H∞ control can guarantee the performance of a nominal ICPT system. However, the dynamic performance may suffer degradation when the parameter deviates from its nominal value. Thus, to enhance the overall dynamic performance of the ICPT system under parameter perturbations, an H∞ control method with a selected weighting function considering parameter perturbations is proposed in this article. First, a generalized state-space model is established to analyze the influence of parameter perturbations on the open-loop system in detail. After that, an uncertain model is established via frequency-domain analysis, generating a numerical solution of a weighting function used to describe the parameter perturbations. Then, the generalized plant with uncertainty is obtained and transferred into a standard H∞ optimal problem via small gain theory. Based on the proposed controller, the dynamic performance and robustness of the closed-loop system are analyzed and verified theoretically. Finally, simulation and experimental results further verify that the proposed H∞ control method can make the closed-loop system obtain good dynamic performance and strong robustness against parameter perturbations. The settling time is less than 6.8 ms without any overshoot when the system suffers parameter perturbations. Besides, the controller can maintain good dynamic performance in the startup and reference tracking experiments. [ABSTRACT FROM AUTHOR]

Despite pedestrian detectors having made significant progress by introducing convolutional neural networks, their performance still suffers degradation, especially in occlusion scenes with more false positives (FPs) and false negatives (FNs). To alleviate the problem, we propose a novel Sequential Attention-based Distinct Part Modeling (SA-DPM) for balanced pedestrian detection. It takes one step further in constructing more robust representation that supports detection with fewer FNs and FPs. Specifically, the Sequential Attention serves as one internal perception process that captures several distinct part areas step by step from each pedestrian proposal (full-body). Different from the previous either-or feature selection, the following Joint Learning attempts to seek a reasonable trade-off between part and full-body features, and combines both features for more accurate classification and regression. Evaluation on the widely used pedestrian datasets including Caltech and Citypersons shows that the proposed SA-DPM achieves promising performance for both non-occluded and occluded pedestrian detection tasks, especially on Caltech Heavy Occlusion set, which yields a new state-of-the-art miss rate by 30.18% and outperforms the second best detector by 6.32%. [ABSTRACT FROM AUTHOR]

Remote sensing images obtained from onboard linear array cameras suffer from geometric disturbances in the presence of attitude jitter from a satellite platform. Thus, platform jitter estimation is essential for improving satellite stability and enhancing remote sensing image quality. In this article, a novel integration framework is designed to estimate attitude fluctuation, which combines multispectral advanced spaceborne thermal emission and reflection/short-wave-infrared (ASTER/SWIR) data and blurred star images from onboard optical sensors. First, a multilevel XGBoost algorithm is applied to learn the nonlinear relationship between the star blurring and instantaneous jitter displacement. An image registration method and a deconvolutional process are then introduced to remove the satellite jitter from ASTER/SWIR data. Next, an augmented $H$$_\infty$ filter is proposed to fuse and estimate the satellite jitter from two algorithms. Simulating experiment results indicate that the jitter estimation error of the proposed integration framework is reduced by 40%. Compared with the existing jitter estimation methods only based on remote sensing image processing, our strategy has better robustness and accuracy, especially on extreme ground scenes. [ABSTRACT FROM AUTHOR]

In this article, we propose a novel tensorial approach, namely, generalized tensor regression, for hyperspectral image classification. First, a simple and effective classifier, i.e., the ridge regression for multivariate labels, is extended to its tensorial version by taking advantages of tensorial representation. Then, the discrimination information of different modes is exploited to further strengthen the capacity of the model. Moreover, the model can be simplified and solved easily. Different from traditional tensorial methods, the proposed model can be utilized to capture not only the intrinsic structure of data in a physical sense but also the generalized relationship of data in a logical sense. Our proposed approach is shown to be effective for different classification purposes on a series of instantiations. Specifically, our experiment results with hyperspectral images collected by the airborne visible/infrared imaging spectrometer, the reflective optics spectrographic imaging system and the ITRES CASI-1500 demonstrate the effectiveness of the proposed approach as compared to other tensor-based classifiers and multiple kernel learning methods. [ABSTRACT FROM AUTHOR]

Recently, a new structure of low density parity check (LDPC) code named raptor-like LDPC code has attracted much attention. It has better performance at low code rate. In this paper, a novel decoding scheme for raptor-like LDPC code is proposed. First, the Gaussian approximation density evolution (GADE) algorithm is used to track and analyze the message transmission during the decoding process. It is found that certain log likelihood ratio (LLR) messages can be approximated by “zero” setting in the early iteration of raptor-like LDPC decoding. In other words, some column and row operations could be eliminated without compromise the performance. Next, we propose a new decoding scheme, which can skip unnecessary column and row operations. In comparison with the traditional belief propagation (BP)-based LDPC decoding scheme, the proposed scheme can reduce the decoding complexity. Additionally, a new algorithm is developed to facilitate the selection of the early iteration number. With this novel design, the proposed decoding scheme performs almost the same as the traditional BP-based scheme. To proof the concept, the raptor-like LDPC codes in the ATSC3.0 digital TV system are used to evaluate the proposed scheme, in comparison with the traditional BP-based schemes, i.e., sum-product algorithm (SPA), offset min-sum algorithm (OMSA), and normalized min-sum algorithm (NMSA). The simulation results confirm that the proposed scheme can reduce the decoding complexity without sacrifice in performance for all SPA, OMSA, and NMSA methods. About 10% complexity reduction can be achieved. This will reduce the buttery consumption for Internet of Things (IoT) and handheld devices. [ABSTRACT FROM AUTHOR]

The traditional degree distribution framework is a useful tool to analyze irregular low density parity check (LDPC) code. However, it cannot be used to analyze raptor-like LDPC code as it cannot describe the most important structural feature of raptor-like LDPC code, which is that any check node is connected to at most one degree-one variable node. In this paper, we propose a two-type degree distribution framework for raptor-like LDPC code. The proposed framework can describe the special structural feature of raptor-like LDPC code by dividing the check nodes into two different sets. One is the set of check nodes connected to those degree-one variable nodes, the other is the set of check nodes unconnected to those degree-one variable nodes. Furthermore, we propose a two-type discrete density evolution algorithm and a two-type Gaussian approximation density evolution algorithm based on the proposed two-type degree distribution framework to estimate the threshold of raptor-like LDPC code. During the decoding analysis, the proposed algorithms treat the check nodes belonging the two separate sets differently. Compared with the traditional multi-edge type (MET) algorithm, the proposed algorithms allow us to calculate the threshold more quickly without much sacrifice in accuracy. For some raptor-like LDPC codes adopted by the Advanced Television Systems Committee 3.0 (ATSC3.0) standard, the thresholds estimated by the proposed algorithms can be very close to the thresholds estimated by the traditional MET algorithm. The deviations are less than 0.07 dB. Meanwhile, the proposed algorithms can tremendously reduce the computation complexity because of the two-type degree distribution framework and the Gaussian approximation. [ABSTRACT FROM AUTHOR]

Cloud and cloud shadow detection in remote sensing imagery is important for its wide range of applications. Traditionally, the detection is usually based on the manually designed thresholds from multiband, which is complicated and of multistage. To simplify the process of cloud and cloud shadow detection and improve the performance, we propose a multilevel feature fused segmentation network (MFFSNet), which can be trained end-to-end without any hand-tuned parameters. Specifically, a fully convolutional network is proposed for cloud and cloud shadow features learning. Then, we utilize a novel pyramid pooling module to extract contextual relation between cloud and shadow. Furthermore, a special multilevel feature fused structure is designed to combine semantic information with spatial information from different levels, so that we can better handle the multiscale objects and produce detailed segmentation boundaries. Experiments show that the MFFSNet outperforms the state-of-the-art methods and achieves high accuracies of 98.69% and 98.92% for cloud and cloud shadow detection. [ABSTRACT FROM AUTHOR]

Linearizability is an important correctness criterion for implementations of concurrent objects. Automatic checking of linearizability is challenging because it requires checking that: 1) All executions of concurrent operations are serializable, and 2) the serialized executions are correct with respect to the sequential semantics. In this work, we describe a method to automatically check linearizability based on refinement relations from abstract specifications to concrete implementations. The method does not require that linearization points in the implementations be given, which is often difficult or impossible. However, the method takes advantage of linearization points if they are given. The method is based on refinement checking of finite-state systems specified as concurrent processes with shared variables. To tackle state space explosion, we develop and apply symmetry reduction, dynamic partial order reduction, and a combination of both for refinement checking. We have built the method into the PAT model checker, and used PAT to automatically check a variety of implementations of concurrent objects, including the first algorithm for scalable nonzero indicators. Our system is able to find all known and injected bugs in these implementations. [ABSTRACT FROM PUBLISHER]

Natural type selection was originally proposed by Zamir and Rose for universal single description coding. In this paper, we generalize this principle to universal multiple description coding (MDC). Two schemes based on random codebooks are proposed: one is of fixed distortion and the other is of fixed weight. Their operational sum-rate-distortion functions are derived, which coincide with the EGC (El Gamal-Cover) sum-rate bound if the parameters of the schemes are optimized. It is also shown that in both schemes the joint type of reconstruction codewords can be used to improve the rate-distortion (R-D) performance. Based on our theoretical results, a practical universal scheme is proposed by leveraging the MDC methods based on low-density generator matrix (LDGM) codes. The performance of this scheme is compared experimentally with the EGC bound, which shows its effectiveness. [ABSTRACT FROM AUTHOR]