检索结果-内蒙古大学图书馆

Progressive Diversified Augmentation for General Robustness of DNNs: A Unified Approach

IEEE TRANSACTIONS ON image processing 2021年 30卷 8955-8967页

作者： Yu, Hang Liu, Aishan Li, Gengchao Yang, Jichen Zhang, Chongzhi Beihang Univ State Key Lab Software Dev Environm Beijing 100191 Peoples R China Univ Southern Calif Dept Comp Sci Los Angeles CA 90007 USA

Adversarial images are imperceptible perturbations to mislead deep neural networks (DNNs), which have attracted great attention in recent years. Although several defense strategies achieved encouraging robustness against adversarial samples, most of them still failed to consider the robustness on common corruptions (e.g. noise, blur, and weather/digital effects). To address this problem, we propose a simple yet effective method, named Progressive Diversified Augmentation (PDA), which improves the robustness of DNNs by progressively injecting diverse adversarial noises during training. In other words, DNNs trained with PDA achieve better general robustness against both adversarial attacks and common corruptions than other strategies. In addition, PDA also enjoys the advantages of spending less training time and keeping high standard accuracy on clean examples. Further, we theoretically prove that PDA can control the perturbation bound and guarantee better robustness. Extensive results on CIFAR-10, SVHN, imageNet, CIFAR-10-C and imageNet-C have demonstrated that PDA comprehensively outperforms its counterparts on the robustness against adversarial examples and common corruptions as well as clean images. More experiments on the frequency-based perturbations and visualized gradients further prove that PDA achieves general robustness and is more aligned with the human visual system.

关键词： Robustness Training Handheld computers Perturbation methods Complexity theory Streaming media Standards Adversarial examples deep learning general robustness computer vision

来源：评论

学校读者我要写书评

暂无评论

Adversarial convolutional network for esophageal tissue segmentation on OCT images

引用

BIOMEDICAL OPTICS EXPRESS 2020年第6期11卷 3095-3110页

作者： Wang, Cong Gan, Meng Zhang, Miao Li, Deyin Chinese Acad Sci Suzhou Inst Biomed Engn & Technol Jiangsu Key Lab Med Opt Suzhou 215163 Peoples R China Soochow Univ Sch Elect & Informat Engn Suzhou 215006 Peoples R China

Automatic segmentation is important for esophageal OCT image processing, which is able to provide tissue characteristics such as shape and thickness for disease diagnosis. Existing automatical segmentation methods based on deep convolutional networks may not generate accurate segmentation results due to limited training set and various layer shapes. This study proposed a novel adversarial convolutional network (ACN) to segment esophageal OCT images using a convolutional network trained by adversarial learning. The proposed framework includes a generator and a discriminator, both with U-Net alike fully convolutional architecture. The discriminator is a hybrid network that discriminates whether the generated results are real and implements pixel classification at the same time. Leveraging on the adversarial training, the discriminator becomes more powerful. In addition, the adversarial loss is able to encode high order relationships of pixels, thus eliminating the requirements of post-processing. Experiments on segmenting esophageal OCT images from guinea pigs confirmed that the ACN outperforms several deep learning frameworks in pixel classification accuracy and improves the segmentation result. The potential clinical application of ACN for detecting eosinophilic esophagitis (EoE), an esophageal disease, is also presented in the experiment. (C) 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement

关键词： image processing Machine vision Medical image processing Optic nerve Optical coherence tomography Segmentation

来源：评论

学校读者我要写书评

暂无评论

A Demographic Vulnerability of Pandemic Prediction: deep learning for Effective Social Distancing Beyond Lockdowns

A Demographic Vulnerability of Pandemic Prediction: Deep Lea...

引用

IEEE International Conference on Signal processing, Communications and Computing (ICSPCC)

作者： Ishfaq Ahmad Asad Ullah Amna Khatoon Li Na School of Marine Science and Technology Northwestern Polytechnical University Xi'an China School of Information Engineering Eurasia University Xi'an China Department of Information Engineering Chang'an University Xi'an China

Pandemic outbreaks are crucial to handle, like the COVID-19 pandemic, spanning from 2020 to 2022 and affecting different countries worldwide, exposing vulnerabilities in social distancing measures during the pandemic's peak. The study presents a fertile, deep learning model designed to enhance future pandemic prevention strategies by recognizing the multifaceted challenges of lockdowns, such as economic adversity, disrupted daily routines, and international trading halts. Our model leverages a comprehensive dataset encompassing demographic information on pandemic dynamics to identify regions and communities susceptible to lapses in social distancing. Crucially, our model goes beyond the constraints of lockdowns by offering real-time monitoring and alerting capabilities. It provides early warnings to health authorities, policymakers, and the public, and these alerts empower decision-makers to implement targeted interventions. This research heralds a transformative approach to pandemic preparedness, harnessing the potential of deep learning to protect vulnerable populations and prevent future outbreaks worldwide.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Label Super-Resolution for 3D Magnetic Resonance images using Deformable U-net

Label Super-Resolution for 3D Magnetic Resonance Images usin...

引用

Conference on Medical Imaging - image processing

作者： Liu, Di Liu, Jiang Liu, Yihao Tao, Ran Prince, Jerry L. Carass, Aaron Beijing Inst Technol Dept Informat & Elect Beijing 100081 Peoples R China Johns Hopkins Univ Dept Elect & Comp Engn Baltimore MD 21218 USA

ISBN: (纸本)9781510640221

Robust and accurate segmentation results from high resolution (HR) 3D Magnetic Resonance (MR) images are desirable in many clinical applications. State-of-the-art deep learning methods for image segmentation require external HR atlas image and label pairs for training. However, the availability of such HR labels is limited due to the annotation accuracy and the time required to manually label. In this paper, we propose a 3D label super resolution (LSR) method which does not use an external image or label from a HR atlas data and can reconstruct HR annotation labels only reliant on a LR image and corresponding label pairs. In our method, we present a Deformable U-net, which uses synthetic data with multiple deformation for training and an iterative topology check during testing, to learn a label slice evolving process. This network requires no external HR data because a deformed version of the input label slice acquired from the LR data itself is used for training. The trained Deformable U-net is then applied to through-plane slices to estimate HR label slices. The estimated HR label slices are further combined by label a fusion method to obtain the 3D HR label. Our results show significant improvement compared to competing methods, in both 2D and 3D scenarios with real data.

关键词： label super resolution MRI deformable model deep learning

来源：评论

学校读者我要写书评

暂无评论

FaceMask: A New image Dataset for the Automated Identification of People Wearing Masks in the Wild

引用

SENSORS 2022年第3期22卷 896-896页

作者： Vrigkas, Michalis Kourfalidou, Evangelia-Andriana Plissiti, Marina E. Nikou, Christophoros Univ Western Macedonia Dept Commun & Digital Media Kastoria 52100 Greece Univ Ioannina Dept Comp Sci & Engn Ioannina 45110 Greece

The rapid spread of the COVID-19 pandemic, in early 2020, has radically changed the lives of people. In our daily routine, the use of a face (surgical) mask is necessary, especially in public places, to prevent the spread of this disease. Furthermore, in crowded indoor areas, the automated recognition of people wearing a mask is a requisite for the assurance of public health. In this direction, image processing techniques, in combination with deep learning, provide effective ways to deal with this problem. However, it is a common phenomenon that well-established datasets containing images of people wearing masks are not publicly available. To overcome this obstacle and to assist the research progress in this field, we present a publicly available annotated image database containing images of people with and without a mask on their faces, in different environments and situations. Moreover, we tested the performance of deep learning detectors in images and videos on this dataset. The training and the evaluation were performed on different versions of the YOLO network using Darknet, which is a state-of-the-art real-time object detection system. Finally, different experiments and evaluations were carried out for each version of YOLO, and the results for each detector are presented.

关键词： face-mask mask detector dataset neural networks Darknet YOLO

来源：评论

学校读者我要写书评

暂无评论

Solving Resource Allocation and Production Scheduling Problems in Aerospace Industry Using Double deep Q-Network with State Feature images

SSRN

引用

SSRN 2024年

作者： Yang, Bing Chen, Xingyan Zhang, Chenyang Xiong, Jian School of Business Administration Southwestern University of Finance and Economics Chengdu611130 China School of Computer and Artificial Intelligence Southwestern University of Finance and Economics Chengdu611130 China School of Data Science City University of Hong Kong Kowloon 999077 Hong Kong

In the aerospace industry, resource allocation and production scheduling are crucial challenges. The demand for top-quality products and streamlined operations has amplified these challenges. While traditional methods have made some progress in addressing these issues, recent research suggests that more flexible and adaptive methods are needed to keep up with the evolving manufacturing environment. This paper presents a novel deep reinforcement learning framework that utilizes a Double deep Q-Network (DDQN) model for resource allocation and production scheduling in the aerospace industry. The complete scheduling process is accurately represented using multi-channel image data and converted into a high-performing multi-stage sequential decision-making model that aligns with customized manufacturing orders. Key information, including processing time, start time, end time, and equipment utilization at each scheduling step, will be fully utilized and adaptable to any non-initial state of the scheduling environment. Additionally, a novel reward function has been developed that minimizes makespan and maximizes resource utilization, leading to increased cumulative rewards and improved overall performance. Experimental results in a real-world production environment indicate the efficacy and potential of the proposed approach for resource allocation and production scheduling in the aerospace industry. © 2024, The Authors. All rights reserved.

关键词： Aerospace industry

来源：评论

学校读者我要写书评

暂无评论

Enhanced Laboratory Safety Education Through Interactive Applications of Machine learning-Boosted image processing Technologies

引用

TRAITEMENT DU SIGNAL 2023年第6期40卷 2623-2633页

作者： Zhang, Lijiao Wu, Jianzhen Wei, Jinxin Yu, Xiaoyan Yu, Jie Yuan, Bo Fuzhou Univ Mech & Elect Engn Practice Ctr Fuzhou 350108 Peoples R China Fujian Vocat Coll Agr Fuzhou 350007 Peoples R China Fuzhou Univ Sch Mech Engn & Automat Fuzhou 350108 Peoples R China Quanzhou Bingdian Technol Co LTD Quanzhou 362000 Peoples R China

Laboratory safety education is fundamental to the smooth conduct of scientific research. Traditional educational models are often limited in interactivity and real-time feedback and they struggle to satisfy the increasing needs of laboratory safety management. This study aims to enhance the interactivity and efficacy of laboratory safety education by adopting machine learning-enhanced image processing technologies. To cope with the noise issue in hazardous behavior data within laboratories, wavelet threshold denoising is applied to significantly improve data usability. To deal with the imbalance in data samples of laboratory hazard behaviors, an adaptive boundary data augmentation algorithm is introduced to balance the dataset and strengthen the model's generalization capability. A breakthrough in extracting spatio-temporal features of hazardous behaviors in laboratories is achieved through an improved Spatio-Temporal Graph Convolutional Network (STGCN) model, enabling effective recognition and classification of hazardous behaviors. The outcome attained in this study has important implications for enhancing the interactivity and practicality of laboratory safety education and it can expand new research directions for the application of machine learning in the field of image processing.

关键词： laboratory safety education machine learning, image processing wavelet threshold denoising data augmentation Spatio-Temporal Graph Convolutional Network (ST-GCN) hazardous behavior recognition

来源：评论

学校读者我要写书评

暂无评论

Interfacing Kera’s deep learning technique for real-time Age and Gender Prediction

Interfacing Kera’s deep learning technique for Real-Time Ag...

引用

International Conference on Computer Communication and Informatics (ICCCI)

作者： Sahiti Vankayalapati R Shekar Y. Rupa Sindhu M. Manvith Reddy P. Sasidhar Reddy Ali Baig Mohammad Dept. of ECE Koneru Lakshmaiah Educational Foundation Guntur India Department of ECE School of Engineering Presidency University Bangalore School of Electronics and Communication Engineering REVA University Bengaluru India

The 21st century has seen the humans achieved many unimaginable achievements in the world of technology. This technology can be used in a different way as individual desire. It can be used to identify a person’s age and also identify the gender of a particular person by detecting face of the person with help of a camera image or even with a real time video. In this research paper, a complete procedure, what are the different methodologies and algorithms that can be used, and which one is more accurate method to be used will be shown. The aim of this work is designing as well developing model which can detect faces thereby predict the age and gender from real time images. This model can be used in police surveillance, policing to dating applications and much more. deep learning techniques are used in designing this model. Convolution neural network and computer vision is also being used in this paper to adapt to real time processing. Convolutional neural network along with the computer vision and keras is used in training the model. After extensive testing and training, accuracy of 85 percent is achieved for the developed model. This model can be broadly used in airports, age restricted areas or buildings and can be used in restricting the alcohol access to the under aged people.

关键词： Training Adaptation models Computational modeling Atmospheric modeling Face recognition Streaming media Predictive models

来源：评论

学校读者我要写书评

暂无评论

learning About the World by learning About images

引用

CURRENT DIRECTIONS IN PSYCHOLOGICAL SCIENCE 2021年第2期30卷 120-128页

作者： Storrs, Katherine R. Fleming, Roland W. Justus Liebig Univ Giessen Dept Expt Psychol Giessen Germany Univ Marburg Ctr Mind Brain & Behav CMBB Marburg Germany Justus Liebig Univ Giessen Giessen Germany

One of the deepest insights in neuroscience is that sensory encoding should take advantage of statistical regularities. Humans' visual experience contains many redundancies: Scenes mostly stay the same from moment to moment, and nearby image locations usually have similar colors. A visual system that knows which regularities shape natural images can exploit them to encode scenes compactly or guess what will happen next. Although these principles have been appreciated for more than 60 years, until recently it has been possible to convert them into explicit models only for the earliest stages of visual processing. But recent advances in unsupervised deep learning have changed that. Neural networks can be taught to compress images or make predictions in space or time. In the process, they learn the statistical regularities that structure images, which in turn often reflect physical objects and processes in the outside world. The astonishing accomplishments of unsupervised deep learning reaffirm the importance of learning statistical regularities for sensory coding and provide a coherent framework for how knowledge of the outside world gets into visual cortex.

关键词： deep learning efficient coding human vision statistical learning unsupervised learning visual perception

来源：评论

学校读者我要写书评

暂无评论

Two-Stream Spatial-Temporal Graph Convolutional Networks for Driver Drowsiness Detection

引用

IEEE TRANSACTIONS ON CYBERNETICS 2022年第12期52卷 13821-13833页

作者： Bai, Jing Yu, Wentao Xiao, Zhu Havyarimana, Vincent Regan, Amelia C. Jiang, Hongbo Jiao, Licheng Xidian Univ Sch Artificial Intelligence Xian 710071 Peoples R China Xidian Univ Key Lab Intelligent Percept & Image Understanding Minist Educ Xian 710071 Peoples R China Hunan Univ Coll Comp Sci & Elect Engn Changsha 410082 Peoples R China Ecole Normale Super Dept Appl Sci Bujumbura 6983 Burundi Univ Calif Irvine Dept Comp Sci & Inst Transportat Studies Irvine CA 92697 USA

Convolutional neural networks (CNNs) have achieved remarkable performance in driver drowsiness detection based on the extraction of deep features of drivers' faces. However, the performance of driver drowsiness detection methods decreases sharply when complications, such as illumination changes in the cab, occlusions and shadows on the driver's face, and variations in the driver's head pose, occur. In addition, current driver drowsiness detection methods are not capable of distinguishing between driver states, such as talking versus yawning or blinking versus closing eyes. Therefore, technical challenges remain in driver drowsiness detection. In this article, we propose a novel and robust two-stream spatial-temporal graph convolutional network (2s-STGCN) for driver drowsiness detection to solve the above-mentioned challenges. To take advantage of the spatial and temporal features of the input data, we use a facial landmark detection method to extract the driver's facial landmarks from real-time videos and then obtain the driver drowsiness detection result by 2s-STGCN. Unlike existing methods, our proposed method uses videos rather than consecutive video frames as processing units. This is the first effort to exploit these processing units in the field of driver drowsiness detection. Moreover, the two-stream framework not only models both the spatial and temporal features but also models both the first-order and second-order information simultaneously, thereby notably improving driver drowsiness detection. Extensive experiments have been performed on the yawn detection dataset (YawDD) and the National TsingHua University drowsy driver detection (NTHU-DDD) dataset. The experimental results validate the feasibility of the proposed method. This method achieves an average accuracy of 93.4% on the YawDD dataset and an average accuracy of 92.7% on the evaluation set of the NTHU-DDD dataset.

关键词： Feature extraction Faces Mouth Brain modeling Vehicles Videos Support vector machines Driver drowsiness detection facial landmark detection graph convolution networks (GCNs)

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：