Hyperspectral unmixing has recently attracted much attention in the field of spectral image analysis. Unsupervised methods based on autoencoder can achieve excellent unmixing performance. However, these methods have unsatisfactory unmixing results for different substances with similar materials in complex scenarios. In this letter, we propose a new dual-branch cross weighting network (DCWNet) for multimodal hyperspectral unmixing. It can not only use the spectral feature information of spectral images but also acquire the spatial feature information of light detection and ranging (LiDAR) data simultaneously. Specifically, we build a spatial channel augmentation (SCA) block to help the network acquire spatial information more accurately from the horizontal, vertical, and channel aspects, respectively. We further construct an adaptive feature selection module for effectively using the spectral feature information and spatial feature information to better focus on the discrimination of materials in the scene through weighting fusion. Experimental results on two real multimodal datasets demonstrate the competitiveness and effectiveness of our proposed DCWNet in comparison to five representative methods.

This article explores the fixed-time tracking control problem for stochastic nonstrict systems with unknown backlash-like hysteresis properties. To this end, a novel criterion of semiglobally practical fixed-time stochastic stability is established and proved. First, a continuous-time dynamic model that can be solved explicitly is constructed to model the discontinuous backlash-like hysteresis nonlinear behavior approximatively. Second, the coupling relationship between stochastic disturbance and hysteresis nonlinearity is analyzed, and the role of the coupling terms is attributed to each subsystem using the inequality expansion and the summation order transformation techniques. Then, based on the above analysis, a memory-free stochastic fixed-time control law without constructing the hysteresis inverse is developed recursively in the framework of backstepping by means of the It$\hat {o}$stochastic differential equation theory and the adaptive fuzzy technique, which can achieve semiglobally practical fixed-time stability of stochastic systems with hysteresis properties and the tracking error can converge to a small neighborhood near the origin. Finally, simulation studies for a numerical simulation example and a cart moving on a plane example are shown to verify the feasibility of the rendered approach.

Generative Adversarial Network (GAN)-based image cartoonization has made great progress. They usually use a “single-encoding adversarial feedback architecture” to generate cartoon image in a similar cartoon-style domain. However, this architecture cannot generate a satisfactory cartoonized image with both high style similarity and visual fidelity. In this work, to relieve this problem, we propose a novel dual-encoding matching adversarial learning dubbed DEMAL for image cartoonlization. Particularly, we first design a dual-encoding matching (DEM) by using a pair of dual encoders and a statistical matching module (SM) to match the content-style feature encodings extracted separately in the statistical space. We then construct double-structure style discriminators to adversarially learn global and local feature representations of cartoon-style via the improved loss function. Furthermore, we also propose a pre-training strategy for the DEMAL to achieve the best FID and ArtFID distance. Extensive experiments have demonstrated that our proposed DEMAL achieves high visual fidelity and style similarity compared to the previous representative baseline cartoonization methods. Code is available athttps://github.com/ZYDeeplearning/DEMAL-Model.

Abstract Semantic segmentation is a fundamental technology for autonomous driving. It has a high demand for inference speed and accuracy. However, a good trade‐off between accuracy and latency is yet not present in existing semantic segmentation approaches. Due to the limitation of speed, the authors cannot increase the number of network layers without limit and cannot design modules like in the networks without real‐time. It is a challenging problem how to design a model with good performance under limited resources. To alleviate these issues, in this study, the authors propose a refinement co‐supervision network (RCNet), which is real‐time on a high‐resolution image (1024×2048). The authors first construct the context refinement module, which can provide low computation cost way for obtaining the large receptive field and context information. Furthermore, a boundary co‐supervision mechanism is proposed. It strengthens the optimisation of easily neglected boundaries and small targets. Experimental results reveal that the proposed RCNet outperforms seven representative semantic segmentation methods.

Salient object detection in optical remote sensing images (RSI-SOD) is a newly developing task and has attracted extensive attention. Although some satisfactory performances have been achieved, there are still some poor performance situations, such as complex backgrounds and objects with various scales. Because the existing methods have the issue of insufficient adjacent feature representation and inappropriate constraint supervision. To alleviate the issues, we attempt to utilize the adjacent complementary and introduce the global constraints. Accordingly, we propose the adjacent complementary network (ACNet) with global constraints for RSI-SOD. Specifically, in the encoder, we mainly design the adjacent complementary module (ACM) to complement the features among the current layer and adjacent layers. Therefore, it can effectively aggregate saliency features. Besides, we have equipped a global average pooling (GAP) to enhance global information. In the decoder, we design a fusion enhancement module (FEM) to enhance the feature conversion between the coding layer and decoding layer. Finally, we employ a novel loss, which introduces global constraints and can supervise the objects with various scales. Extensive studies on two benchmark datasets demonstrate that ACNet is competitive and outperforms the 15 state-of-the-art (SOTA) approaches.

Hyperspectral unmixing is an essential research topic for spectral data analysis due to the existence of mixed pixels. Recently, many methods based on sparse nonnegative matrix factorization (NMF) have been widely used for unmixing by incorporating similarity preserving. However, most of them conduct the similarity learning and unmixing by using two separate steps, which may lead to the case that the learned similarity matrix is not the optimal one for unmixing. Thus, the performance of unmixing would be influenced and become undesirable. To alleviate this problem, we propose an adaptive relationship preserving-based sparse NMF (ARP-NMF) for hyperspectral unmixing. Typically, we regard similarity learning and unmixing as an alternative optimization process. During this process, the learned similarity weights and unmixing results can be mutually improved. Thus, our proposed method can learn the optimal similarity weights for unmixing and obtain better generalization ability for different hyperspectral images than traditional methods. Moreover, by using the spectral and spatial local structure, our ARP-NMF method effectively preserves structure consistency between pixels and abundances. Experimental results both on the synthetic data and the real data reveal that our proposed method outperforms several representative sparse NMF-based methods.

Cartoonization as a special type of artistic style transfer is a difficult image processing task. The current existing artistic style transfer methods cannot generate satisfactory cartoon-style images due to that artistic style images often have delicate strokes and rich hierarchical color changes while cartoon-style images have smooth surfaces without obvious color changes, and sharp edges. To this end, we propose a cartoon loss based generative adversarial network (CartoonLossGAN) for cartoonization. Particularly, we first reuse the encoder part of the discriminator to build a compact generative adversarial network (GAN) based cartoonization architecture. Then we propose a novel cartoon loss function for the architecture. It can imitate the process of sketching to learn the smooth surface of the cartoon image, and imitate the coloring process to learn the coloring of the cartoon image. Furthermore, we also propose an initialization strategy, which is used in the scenario of reusing the discriminator to make our model training easier and more stable. Extensive experimental results demonstrate that our proposed CartoonLossGAN can generate fantastic cartoon-style images, and outperforms four representative methods.

With the development of vehicle electrification, automation and connectivity, collaborative optimization among the traffic throughput, driving comfort, fuel economy and driving safety targets is still a huge challenging barrier for a connected and automated fuel cell/battery hybrid electric vehicle. Hence, this paper proposes an optimal car-following energy management strategy (EMS) that combines energy management and adaptive cruise control considering the above targets. Specifically, based on vehicle-to-vehicle and vehicle-to-infrastructure information, an optimal following distance algorithm is developed to obtain the optimal following distance considering driving safety, driving comfort and traffic throughput. Then, based on the established vehicle longitudinal dynamics model, an adaptive cruise controller using back-stepping technique is designed to accurately track optimal following distance. Meantime, combining the obtained controller, optimal EMS based on equivalent consumption minimization strategy is proposed to coordinate the output power of fuel cell and battery to improve fuel economy. The simulations of short and long-term driving cycles indicate that the proposed method can reduce hydrogen consumption by 12.12%, jerk by 61.21%, and keep the desired following distance tracking error within 0.5m.

Many representation methods were built for gray image textures. However, they are not effective for color textures in general. To alleviate this problem, in this paper we propose a novel Compact Interchannel Sampling Difference Descriptor (CISDD) for color texture classification. In particular, considering sampling-based method can capture more directional information, we first use a heavy-tailed distribution, t-distribution to generate sample points in the image patch to calculate the micro-block difference. Then we model the interchannel relationship of color texture image by using dense micro-block differences. Furthermore, we utilize principal component analysis (PCA) to reduce the dimensions of the features encoded by the Fisher vector, and construct a Compact Interchannel Sampling Difference Descriptor (CISDD) for representing color texture image. Finally, experimental results on five published standard texture datasets (KTH-TIPS, VisTex, CUReT, USPTex and Colored Brodatz) reveal that CISDD is effective and outperforms thirteen representative color texture classification methods.

At present, most saliency detection methods are based on fully convolutional neural networks (FCNs). However, FCNs usually blur the edges of salient objects. Due to that, the multiple convolution and pooling operations of the FCNs will limit the spatial resolution of the feature maps. To alleviate this issue and obtain accurate edges, we propose a hierarchical edge refinement network (HERNet) for accurate saliency detection. In detail, the HERNet is mainly composed of a saliency prediction network and an edge preserving network. Firstly, the saliency prediction network is used to roughly detect the regions of salient objects and is based on a modified U-Net structure. Then, the edge preserving network is used to accurately detect the edges of salient objects, and this network is mainly composed of the atrous spatial pyramid pooling (ASPP) module. Different from the previous indiscriminate supervision strategy, we adopt a new one-to-one hierarchical supervision strategy to supervise the different outputs of the entire network. Experimental results on five traditional benchmark datasets demonstrate that the proposed HERNet performs well when compared with the state-of-the-art methods.

Convolutional Neural Network (CNN) has shown their advantages in salient object detection. CNN can generate great saliency maps because it can obtain high-level semantic information. And the semantic information is usually achieved by stacking multiple convolutional layers and pooling layers. However, multiple pooling operations will reduce the size of the feature map and easily blur the boundary of the salient object. Therefore, such operations are not beneficial to generate great saliency results. To alleviate this issue, we propose a novel edge information-guided hierarchical feature fusion network (HFFNet). Our network fuses features hierarchically and retains accurate semantic information and clear edge information effectively. Specifically, we extract image features from different levels of VGG. Then, we fuse the features hierarchically to generate high-level semantic information and low-level edge information. In order to retain better information at different levels, we adopt a one-to-one hierarchical supervision strategy to supervise the generation of low-level information and high-level information respectively. Finally, we use low-level edge information to guide the saliency map generation, and the edge guidance fusion is able to identify saliency regions effectively. The proposed HFFNet has been extensively evaluated on five traditional benchmark datasets. The experimental results demonstrate that the proposed model is fairly effective in salient object detection compared with 10 state-of-the-art models under different evaluation indicators, and it is superior to most of the comparison models.

A dense micro-block difference (DMD)-based method was proposed for performing texture representation that is a fundamental task of image and video analysis. However, it cannot capture effectively the rotation invariance and multiscale spatial information of textures. To alleviate these problems, in this paper, we propose a multiscale symmetric DMD (MSDMD) method for texture classification. In particular, we first combine K-rotation and Gaussian distribution to analyze the Symmetric DMD in order to capture the rotation invariance of textures. Furthermore, we propose a high-order vector of locally aggregated descriptor called HVLAD by incorporating the second-order and third-order statistics into the original vector of VLAD. To effectively extract the spatial information of textures, we implement the above-mentioned steps in a Gaussian pyramid structure to construct an MSDMD feature and use a support vector machine (SVM) to perform texture classification. The experimental results on five available published texture datasets (KTH-TIPS, CUReT, UIUC, UMD, and KTH-TIPS2-b) reveal that our proposed method is effective when compared with 15 representative texture classification methods.

Subspace clustering is a problem of exploring the low-dimensional subspaces of high-dimensional data. State-of-the-art approaches are designed by following the model of spectral clustering-based method. These methods pay much attention to learn the representation matrix to construct a suitable similarity matrix and overlook the influence of the noise term on subspace clustering. However, the real data are always contaminated by the noise and the noise usually has a complicated statistical distribution. To alleviate this problem, in this paper, we propose a subspace clustering method based on Cauchy loss function (CLF). Particularly, it uses CLF to penalize the noise term for suppressing the large noise mixed in the real data. This is due to that the CLF's influence function has an upper bound that can alleviate the influence of a single sample, especially the sample with a large noise, on estimating the residuals. Furthermore, we theoretically prove the grouping effect of our proposed method, which means that highly correlated data can be grouped together. Finally, experimental results on five real data sets reveal that our proposed method outperforms several representative clustering methods.

In complex scenes, foreground saliency can hardly be detected completely, which may further result in the ambiguous cues of objects for other computer vision tasks. In this letter, an extended locality-constrained linear self-coding (eLLsC) scheme is proposed to assist to solve the saliency detection problem under the complex scenes. The locality of both spatial relation and feature distance is preserved in eLLsC, thus making the transformed code involved in the manifold ranking to prompt the generation of the saliency map with more complete foreground and clearer boundary. Experimental results on three saliency detection benchmarks demonstrate the effectiveness of the proposed hybrid method.