检索结果-内蒙古大学图书馆

arXiv 2019年

作者： Hu, Xiaowei Jiang, Yitong Fu, Chi-Wing Heng, Pheng-Ann Department of Computer Science and Engineering Chinese University of Hong Kong Guangdong Provincial Key Laboratory of Computer Vision and Virtual Reality Technology Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences China

This paper presents a new method for shadow removal using unpaired data, enabling us to avoid tedious annotations and obtain more diverse training samples. However, directly employing adversarial learning and cycle-consistency constraints is insufficient to learn the underlying relationship between the shadow and shadow-free domains, since the mapping between shadow and shadow-free images is not simply one-to-one. To address the problem, we formulate Mask-ShadowGAN, a new deep framework that automatically learns to produce a shadow mask from the input shadow image and then takes the mask to guide the shadow generation via re-formulated cycle-consistency constraints. Particularly, the framework simultaneously learns to produce shadow masks and learns to remove shadows, to maximize the overall performance. Also, we prepared an unpaired dataset for shadow removal and demonstrated the effectiveness of Mask-ShadowGAN on various experiments, even it was trained on unpaired data. Copyright © 2019, The Authors. All rights reserved.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

Attention-guided low-light image enhancement

arXiv

引用

arXiv 2019年

作者： Lv, Feifan Lu, Feng State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and Engineering Beihang University Beijing100191 China Peng Cheng Laboratory Shenzhen518000 China

Low-light image enhancement is a challenging task since various factors, including brightness, contrast, artifacts and noise, should be handled simultaneously and effectively. To address such a difficult problem, this paper proposes a novel attention-guided enhancement solution and delivers the corresponding end-to-end multi-branch CNNs. The key of our method is the computation of two attention maps to guide the exposure enhancement and denoising respectively. In particular, the first attention map distinguishes underexposed regions from normally exposed regions, while the second attention map distinguishes noises from real-world textures. Under their guidance, the proposed multi-branch enhancement network can work in an adaptive way. Other contributions of this paper include the " decomposition/multi-branch-enhancement/fusion" design of the enhancement network, the reinforcement-net for contrast enhancement, and the proposed large-scale low-light enhancement dataset. We evaluate the proposed method through extensive experiments, and the results demonstrate that our solution outperforms state-of-the-art methods by a large margin. We additionally show that our method is flexible and effective for other image processing tasks. Copyright © 2019, The Authors. All rights reserved.

关键词： Image enhancement

来源：评论

学校读者我要写书评

暂无评论

Background segmentation for vehicle re-identification

arXiv

引用

arXiv 2019年

作者： Wu, Mingjie Zhang, Yongfei Zhang, Tianyu Zhang, Wenqi Beijing Key Laboratory of Digital Media School of Computer Science and Engineering Beihang University Beijing100191 China State Key Laboratory of Virtual Reality Technology and Systems Beihang University Beijing100191 China

Vehicle re-identification (Re-ID) is very important in intelligent transportation and video surveillance. Prior works focus on extracting discriminative features from visual appearance of vehicles or using visual-spatio-temporal information. However, background interference in vehicle re-identification have not been explored. In the actual large-scale spatio-temporal scenes, the same vehicle usually appears in different backgrounds while different vehicles might appear in the same background, which will seriously affect the re-identification performance. To the best of our knowledge, this paper is the first to consider the background interference problem in vehicle re-identification. We construct a vehicle segmentation dataset and develop a vehicle Re-ID framework with a background interference removal (BIR) mechanism to improve the vehicle Re-ID performance as well as robustness against complex background in large-scale spatio-temporal scenes. Extensive experiments demonstrate the effectiveness of our proposed framework, with an average 9% gain on mAP over state-of-the-art vehicle Re-ID algorithms. Copyright © 2019, The Authors. All rights reserved.

关键词： Security systems

来源：评论

学校读者我要写书评

暂无评论

A Novel Joint Optimization Method for Adaptive Hand-held Video Stabilization Based on Spatialtemporal Consistency

A Novel Joint Optimization Method for Adaptive Hand-held Vid...

引用

2019 International Conference on Modeling, Simulation, Optimization and Numerical Techniques(SMONT 2019)

作者： Xiao Li Shuai Li Hong Qin Aimin Hao State Key Laboratory of Virtual Reality Technology and Systems Beihang University Beihang University Qingdao Research Institute Department of Computer Science Stony Brook University

ISBN: (纸本)9781510885455

Video stabilization for hand-held camera is vital in many high-level video enhancement applications. Although quite a few proposed approaches have achieved remarkable success in recent years, many technical challenges still prevail for video(from hand-held camera) containing wide-range scenes and highlyvariable objects. In particular, we are lacking effective and versatile strategies to adaptively handle saliency preservation, parallax diminution, self-adaptive smoothing, cropping area decrement and video completeness in a consistent spatial-temporal fashion. To ameliorate, this paper develops a novel, joint optimization method to successively respect spatial structure consistency and temporal feature constraints. Our aim is to devise a new adaptive video stabilization technique by resorting to new modeling strategies. This paper's key originality is hinged upon joint utility of both self-adaptive intrinsic mode functions(IMFs) based on empirical mode decomposition(EMD) in temporal domain for video signal and mesh-structure constraint enforcement in the spatial domain. As a result, our new approach can optimize camera trajectory of wobbly video, and synchronously fine-tune the camera path based on key features to make the shaking video much closer to the original trend. To validate our joint optimization approach for adaptive video stabilization, we conduct comprehensive experiments on public benchmarks, and make extensive and quantitative evaluations with available state-of-the-art methods as well as popular commercial software. All of our experiments demonstrate the advantages of the joint optimization method in terms of versatility, accuracy, and efficiency.

关键词： hand-held Camera adaptive video stabilization spatial structure consistency feature-centric EMD self-adaptive IMF selection video extrapolation and interpolation

来源：评论

学校读者我要写书评

暂无评论

Temporal-Spatial Consistency and Sampling Neural Network Based Adaptive Hand-held Video Deblurring

Temporal-Spatial Consistency and Sampling Neural Network Bas...

引用

作者： Xiao LI Shuai LI Hong QIN Ai-min HAO State Key Laboratory of Virtual Reality Technology and Systems Beihang University Beihang University Qingdao Research Institute Department of Computer Science Stony Brook University

Video deblurring for hand-held camera is vital in many high-level video enhancement applications. Although quite a few proposed approaches have achieved remarkable success in recent years, many technical challenges still prevail for video(from hand-held camera) containing wide-range scenes and highly-variable objects. In particular, we are lacking effective and versatile strategies to adaptively handle high-level alignment, parallax diminution, unclear area detection, sharpness enhancement in a consistent fashion. To ameliorate, this paper develops a novel method to successively respect temporal-spatial alignment and precise deblurring. Our aim is to devise a new adaptive video deblurring technique by resorting to new modeling strategies. This paper's key originality is hinged upon the joint utility of both self-adaptive intrinsic mode functions(IMFs) based on empirical mode decomposition(EMD) in the temporal domain for video signal and mesh-structure constraint enforcement in the spatial domain, beside, this paper innovatively jointly utilities the subsampling convolution, upsampling convolution and luckiness threshold value based back propagation, which takes video deblurring as a high-precision adaptive convolution. As a result, our new approach can adaptively improve the clarity of the video, and synchronously optimize the camera trajectory of wobbly video. To validate our optimization approach for adaptive video deblurring, we conduct comprehensive experiments on public benchmarks, and make extensive and quantitative evaluations with available state-of-the-art methods as well as popular commercial software. All of our experiments demonstrate the advantages of the optimization method in terms of versatility, accuracy, and efficiency.

关键词： Neural network Adaptive video deblurring Temporal alignment Spatial consistency alignment Luckiness threshold value

来源：评论

学校读者我要写书评

暂无评论

A Fast Method for Large-Scale Scene Data Acquisition and 3D Reconstruction

A Fast Method for Large-Scale Scene Data Acquisition and 3D ...

引用

IEEE International Symposium on Mixed and Augmented Reality Workshops (ISMARW)

作者： Yao Li Yang Xie Xijing Wang Xun Luo Yue Qi State Key Laboratory of Virtual Reality Technology and Systems Beijing School of Computer Science and Engineering Tianjin University of Technology Tianjin Peng Cheng Laboratory Shenzhen Qingdao Research Institute of Beihang University Qingdao

ISBN: (纸本)9781728147666

Large-scale 3D scene has an important application prospect in virtual reality and augmented reality. However, due to the doubling of the amount of data in the 3D reconstruction of large-scale outdoor scene, the index of time becomes a great challenge under the condition of maintaining a certain degree of accuracy. In this paper, we provide a set of methods of aerial photography data acquisition and cluster-based 3D model reconstruction for large-scale scene. Firstly, for the data acquisition end, we adopt two strategies of track pre-planning and feedback-based trajectory planning to meet the requirements of efficient data acquisition. Secondly, we have designed and implemented a distributed 3D reconstruction system to process a large number of aerial photographs, which can quickly and robustly reconstruct large-scale 3D models. Finally, we simulate the crowd behavior based on the PEM model in the reconstructed 3D scene, which has a useful guiding significance for people's daily activities and emergency problems.

关键词： Three-dimensional displays Image reconstruction Solid modeling Computational modeling Data models Data acquisition Drones

来源：评论

学校读者我要写书评

暂无评论

3D shape co-segmentation via sparse and low rank representations

引用

Science China(Information Sciences) 2018年第5期61卷 285-287页

作者： Liyuan YIN Kan GUO Bin ZHOU Qinping ZHAO State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and EngineeringBeihang University

As a traditional and fundamental research area in computer graphics,3D shape segmentation has provided large support to many other 3D processing technologies.3D shape co-segmentation,which together with model sets,co-segments multiple 3D shape parts among the same category,has been attracting significant attention to fully utilize the common features of model sets.

关键词： NSE shape co-segmentation via sparse and low rank representations PSB

来源：评论

学校读者我要写书评

暂无评论

Outlier detection based on cluster outlier factor and mutual density 10th

Outlier detection based on cluster outlier factor and mutual...

引用

10th International Symposium on Intelligence Computation and Applications, ISICA 2018

作者： Zhang, Zhongping Zhu, Mengfan Qiu, Jingyang Liu, Cong Zhang, Debin Qi, Jie School of Information Science and Engineering Yanshan University QinhuangdaoHebei066004 China The Key Laboratory for Computer Virtual Technology and System Integration of Hebei Province QinhuangdaoHebei066004 China Hebei Education Examinations Authority ShijiazhuangHebei050000 China The First Middle School of Qian An Country Qian’anJilin131400 China

ISBN: (纸本)9789811364723

Outlier detection is an important task in data mining with numerous applications. Recent years, the study on outlier detection is very active, many algorithms were proposed including that based on clustering. However, most outlier detection algorithms based on clustering often need parameters, and it is very difficult to select a suitable parameter for different data set. In order to solve this problem, an outlier detection algorithm called outlier detection based on cluster outlier factor and mutual density is proposed in this paper which combining the natural neighbor search algorithm of the Natural Outlier Factor (NOF) algorithm and based on the Density and Distance Cluster (DDC) algorithm. The mutual density and γ density is used to construct decision graph. The data points with γ density anomalously large in decision graph are treated as cluster centers. This algorithm detect the boundary of outlier cluster using cluster outlier factor called Cluster Outlier Factor (COF), it can automatic find the parameter. This method can achieve good performance in clustering and outlier detection which be shown in the experiments. © Springer Nature Singapore Pte Ltd. 2019.

关键词： Data mining

来源：评论

学校读者我要写书评

暂无评论

Salient object detection with purificatory mechanism and structural similarity loss

arXiv

引用

arXiv 2019年

作者： Li, Jia Su, Jinming Xia, Changqun Ma, Mingcan Tian, Yonghong The State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and Engineering Beihang University Beijing100191 China The Department of Computer Science and Technology Peking University Beijing100871 China Peng Cheng Laboratory Shenzhen518000 China

Image-based salient object detection has made great progress over the past decades, especially after the revival of deep neural networks. By the aid of attention mechanisms to weight the image features adaptively, recent advanced deep learning-based models encourage the predicted results to approximate the ground-truth masks with as large predictable areas as possible, thus achieving the state-of-the-art performance. However, these methods do not pay enough attention to small areas prone to misprediction. In this way, it is still tough to accurately locate salient objects due to the existence of regions with indistinguishable foreground and background and regions with complex or fine structures. To address these problems, we propose a novel convolutional neural network with purificatory mechanism and structural similarity loss. Specifically, in order to better locate preliminary salient objects, we first introduce the promotion attention, which is based on spatial and channel attention mechanisms to promote attention to salient regions. Subsequently, for the purpose of restoring the indistinguishable regions that can be regarded as error-prone regions of one model, we propose the rectification attention, which is learned from the areas of wrong prediction and guide the network to focus on error-prone regions thus rectifying errors. Through these two attentions, we use the Purificatory Mechanism to impose strict weights with different regions of the whole salient objects and purify results from hard-to-distinguish regions, thus accurately predicting the locations and details of salient objects. In addition to paying different attention to these hard-to-distinguish regions, we also consider the structural constraints on complex regions and propose the Structural Similarity Loss. The proposed loss models the region-level pair-wise relationship between regions to assist these regions to calibrate their own saliency values. In experiments, the proposed purificatory mecha

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Mobile Internet Mobile Agent System Dynamic Trust Model for Cloud Computing 1

引用

4th International Conference on Human Centered Computing, HCC 2018

作者： Jiang, Weijin Jiang, Yirong Wang, Yang Chen, Jiahui Xu, Yuhui Tan, Li'na Institute of Big Data and Internet Innovation Mobile E-business Collaborative Innovation Center of Hunan Province Hunan University of Commerce Changsha410205 China Key Laboratory of Hunan Province for New Retail Virtual Reality Technology Hunan University of Commerce Changsha410205 China School of Computer Science and Technology Wuhan University of Technology Wuhan430073 China Tonghua Normal University Tonghua134002 China

ISBN: (数字)9783030151270

ISBN: (纸本)9783030151263

Aiming at the security and trust management of mobile agent system in cloud computing environment, based on the simple public key infrastructure (SPKI) trust mechanism, a mobile agent system objective trust peer management model (MAOTM) is proposed to solve the problem of identity authentication, operation authorization and access control in the mobile agent system;On this basis, the Human Trust Mechanism (HTM) is used to study the subjective trust formation, trust propagation and trust evolution law, and the subjective trust dynamic management algorithm (MASTM) is proposed. Based on the interaction experience between the mobile agent and the execution host and the third-party recommendation information to collect the basic trust data, the public trust host selection algorithm is given. The isolated malicious host algorithm and the integrated trust degree calculation algorithm realize the function of selecting the trusted cluster and isolating the malicious host, so as to enhance the security interaction between the mobile agent and the host. The simulations of the proposed algorithms are verified and proved to be feasible and effective. © 2019, Springer Nature Switzerland AG.

关键词： Mobile agents

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：