检索结果-内蒙古大学图书馆

19th IEEE International Conference on Trust, Security and Privacy in Computing and Communications (IEEE TrustCom)

作者： Li, Meng Zhong, Qi Zhang, Leo Yu Du, Yajuan Zhang, Jun Xiang, Yong Wuhan Univ Technol Wuhan Peoples R China Deakin Univ Geelong Vic Australia Swinburne Univ Technol Melbourne Vic Australia

ISBN: (纸本)9781665403924

Similar to other digital assets, deep neural network (DNN) models could suffer from piracy threat initiated by insider and/or outsider adversaries due to their inherent commercial value. DNN watermarking is a promising technique to mitigate this threat to intellectual property. This work focuses on black-box DNN watermarking, with which an owner can only verify his ownership by issuing special trigger queries to a remote suspicious model. However, informed attackers, who are aware of the watermark and somehow obtain the triggers, could forge fake triggers to claim their ownerships since the poor robustness of triggers and the lack of correlation between the model and the owner identity. This consideration calls for new watermarking methods that can achieve better trade-off for addressing the discrepancy. In this paper, we exploit frequency domain image watermarking to generate triggers and build our DNN watermarking algorithm accordingly. Since watermarking in the frequency domain is high concealment and robust to signal processing operation, the proposed algorithm is superior to existing schemes in resisting fraudulent claim attack. Besides, extensive experimental results on 3 datasets and 8 neural networks demonstrate that the proposed DNN watermarking algorithm achieves similar performance on functionality metrics and better performance on security metrics when compared with existing algorithms.

关键词： deep neural networks frequency transform intellectual property watermarking

来源：评论

学校读者我要写书评

暂无评论

MULTI-SCALE DILATED RESIDUAL CONVOLUTIONAL neural NETWORK FOR HYPERSPECTRAL image CLASSIFICATION 10

MULTI-SCALE DILATED RESIDUAL CONVOLUTIONAL NEURAL NETWORK FO...

引用

10th Workshop on Hyperspectral Imaging and signal processing: Evolution in Remote Sensing (WHISPERS)

作者： Pooja, Kumari Nidamanuri, Rama Rao Mishra, Deepak Indian Inst Space Sci & Technol Dept Avion Thiruvananthapuram 695547 Kerala India Indian Inst Space Sci & Technol Dept Earth & Space Sci Thiruvananthapuram 695547 Kerala India

ISBN: (纸本)9781728152943

Recently, deep Convolutional neural Networks (CNNs) have been extensively studied for hyperspectral image classification. It has undergone significant improvement as compared to conventional classification methods. Yet, there are not much studies have been taken on sub-sampled ground truth dataset in CNN. This paper exploits CNN-based method along with multi-scale and dilated convolution with residual connection concepts for hyperspectral image classification on exclusive real time data set. Two raw and one standard full ground truth Pavia University datasets are used to characterize the performance. Out of raw exclusive datasets, one was taken over urban areas of Ahmedabad, India under ISRO-NASA joint initiative for HYperSpectral Imaging (HYSI) programme, and the other was collected using Hypersec VNIR integrated camera of our institute surroundings from the rooftop of the building.

关键词： Hyperspectral Imaging Deep Learning Convolutional neural Network Multi-scale Dilated Residual Learning

来源：评论

学校读者我要写书评

暂无评论

Improvement of edge-tracking methods using Genetic algorithm and neural network 5

Improvement of edge-tracking methods using Genetic algorithm...

引用

5th Iranian Conference on signal processing and Intelligent Systems (ICSPIS)

作者： Shabankareh, Sajjad Ghazanfari Shabankareh, Saeid Ghazanfari Islamic Azad Univ Fac Elect Engn Shiraz Iran Aix Marseille Univ Fac Sci & Technol Marseille France

ISBN: (纸本)9781728153506

One of the most basic and important operations in the field of image processing is image extraction and detection. Edge recognition is very important for image clarity and image segmentation. The importance of edge detection is that the human eye can recognizes the existence of different objects by observing the edges. So it makes sense to do edge detection before interpreting images in automated systems. Edges are points in the image where the two pixels are two different values, or two values very large in numerical value. It can be said that one of the goals of edge reduction in data size in images is to preserve the original structure and shape of the images. Editing has various applications, including object recognition, segmentation and image coding, in a variety of medical images. One of the problems we encounter when editing images has noises. In this paper, a combination of several standard edge- matching algorithms, neural network and multi-objective genetic algorithm Non-Dominated Sorting Genetic Algorithm (NSGA-II) is used to edge detection. In the proposed method, we give the standard multi-edge finder to a forward propagating(FP) neural network as input. A multi-objective genetic algorithm with Non-Dominated Sorting Genetic Algorithm (NSGA-II was used to select the number of edge detectors to access the neural network and to select the most accurate ones. The genetic algorithm selects the least accurate number of edge detectors to enter the neural network. To evaluate the proposed method, TP, TN, FP, FN criteria were used to compare with other methods. Finally, by comparing the results obtained on different images using existing methods and the proposed method it is observed that the proposed method has better accuracy in detecting the edge of noise images.

关键词： Edge detection neural network multi-objective genetic algorithm with Non-Dominated Sorting Genetic Algorithm (NSGA-II) image processing

来源：评论

学校读者我要写书评

暂无评论

Classification of brain activities during language and music perception

引用

signal image AND VIDEO processing 2019年第8期13卷 1559-1567页

作者： Besedova, Petra Vysata, Oldrich Mazurova, Radka Kopal, Jakub Ondrakova, Jana Valis, Martin Prochazka, Ales Univ Hradec Kralove Dept German Language & Literature Fac Educ Hradec Kralove 50003 Czech Republic Charles Univ Prague Dept Neurol Fac Med Hradec Kralove Hradec Kralove 50005 Czech Republic Univ Chem & Technol Prague Dept Comp & Control Engn Prague 16628 6 Czech Republic Czech Tech Univ Czech Inst Informat Robot & Cybernet Prague 16636 6 Czech Republic

Analysis of brain activities in language perception for individuals with different musical backgrounds can be based upon the study of multichannel electroencephalograhy (EEG) signals acquired in different external conditions. The present paper is devoted to the study of the relationship of mental processes and the perception of external stimuli related to the previous musical education. The experimental set under study included 38 individuals who were observed during perception of music and during listening to foreign languages in four stages, each of which was 5 min long. The proposed methodology is based on the application of digital signal processing methods, signal filtering, statistical methods for signal segment selection and active electrode detection. neural networks and support vector machine (SVM) models are then used to classify the selected groups of linguists to groups with and without a previous musical education. Our results include mean classification accuracies of 82.9% and 82.4% (with the mean cross-validation errors of 0.21 and 0.22, respectively) for perception of language or music and features based upon EEG power in the beta and gamma EEG frequency bands using neural network and SVM classification models.

关键词： Multichannel signal analysis Computational intelligence Cognitive science Linguistics Machine learning

来源：评论

学校读者我要写书评

暂无评论

Deep learning-based embedded license plate localisation system

引用

IET INTELLIGENT TRANSPORT SYSTEMS 2019年第10期13卷 1569-1578页

作者： Yepez, Juan Castro-Zunti, Riel D. Ko, Seok-Bum Univ Saskatchewan Dept Elect & Comp Engn Saskatoon SK Canada

In this study the authors propose novel neural network architecture for license plate localisation (LPL) based on an inverted residual structure where the shortcut connections are between the linear bottleneck layers. This residual structure is used for feature extraction in a modified single shot detector for object detection, where standard convolutions are replaced with depthwise separable convolutions in classification layers. The proposed deep learning (DL) solution was tested against three popular international research databases and achieves state-of-the-art results, proving that the proposed model is accurate and robust. Across those databases, the proposed model surpasses other recent LPL works, including DL-based methods, in terms of accuracy and speed. The authors show the proposed architecture to have significant speedup and computational efficiency gains over other DL models, and to have fast per-image localisation processing times sufficient for applications deployed on expensive and commodity hardware alike. Using a novel multi-threading video capture with motion detection then inference algorithm, the authors increase computational efficiency and drop fewer frames overall, allowing for increased performance. Repeated tests show that the proposed method is well-suited to real-time and highly accurate LPL, regardless of hardware.

关键词： video signal processing traffic engineering computing image classification learning (artificial intelligence) object detection multi-threading feature extraction inference mechanisms image capture image motion analysis neural net architecture neural network architecture inverted residual structure linear bottleneck layers feature extraction modified single shot detector object detection depthwise separable convolutions classification layers fast per-image localisation motion detection deep learning multithreading video capture embedded license plate localisation system inference algorithm

来源：评论

学校读者我要写书评

暂无评论

Human Action Recognition using Pre-trained Convolutional neural Networks 20

Human Action Recognition using Pre-trained Convolutional Neu...

引用

2020 2nd International Conference on Video, signal and image processing

作者： Cherry Aly Fazly Salleh Abas Hock Ann Goh Multimedia University Malaysia

ISBN: (纸本)9781450388931

Recognition of human action is one of the challenges in the field of artificial intelligence. Deep learning model has become a research issue in action recognition applications due to its ability to outperform traditional machine learning approaches. The Convolutional neural Network is one of the architectures commonly used in most action recognition works. There are different models in the Convolutional neural Network, but no study has been done to evaluate which model has the best performance in understanding human actions. Thus, in this paper, we compare the performance of two separate pre-trained models of deep Convolutional neural Network in classifying the human actions to identify the different behaviours. GoogleNet and AlexNet are the used two models with fine-tuned parameters used for comparison, in addition, to use Long-Short Term Memory for the video's labels prediction. The paper's main contribution is that it offers a performance analysis of two separate fine-tuned deep CNN pre-trained models compared to the results of other recently proposed human action recognition methods applied on KTH, Weizmann, UCF11(YouTube actions) and UCF-Sports datasets.

关键词： Action recognition AlexNet Convolutional neural network (CNN) Deep learning GoogleNet Long-Short Term Memory (LSTM)

来源：评论

学校读者我要写书评

暂无评论

Distribution mismatch correction for improved robustness in deep neural networks

arXiv

引用

arXiv 2021年

作者： Fuchs, Alexander Knoll, Christian Pernkopf, Franz Signal Processing and Speech Communication Laboratory Graz University of Technology

Deep neural networks rely heavily on normalization methods to improve their performance and learning behavior. Although normalization methods spurred the development of increasingly deep and efficient architectures, they also increase the vulnerability with respect to noise and input corruptions. In most applications, however, noise is ubiquitous and diverse;this can often lead to complete failure of machine learning systems as they fail to cope with mismatches between the input distribution during training- and test-time. The most common normalization method, batch normalization, reduces the distribution shift during training but is agnostic to changes in the input distribution during test time. This makes batch normalization prone to performance degradation whenever noise is present during test-time. Sample-based normalization methods can correct linear transformations of the activation distribution but cannot mitigate changes in the distribution shape;this makes the network vulnerable to distribution changes that cannot be reflected in the normalization parameters. We propose an unsupervised non-parametric distribution correction method that adapts the activation distribution of each layer. This reduces the mismatch between the training and test-time distribution by minimizing the 1-D Wasserstein distance. In our experiments, we empirically show that the proposed method effectively reduces the impact of intense image corruptions and thus improves the classification performance without the need for retraining or fine-tuning the model. Copyright © 2021, The Authors. All rights reserved.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Label-free optical hemogram of granulocytes enhanced by artificial neural networks

引用

OPTICS EXPRESS 2019年第10期27卷 13706-13720页

作者： Gupta, Roopam K. Chen, Mingzhou Malcolm, Graeme P. A. Hempler, Nils Dholakia, Kishan Powis, Simon J. Univ St Andrews Sch Med Biomed Sci Res Complex St Andrews KY16 9TF Fife Scotland Univ St Andrews Sch Phys & Astron SUPA St Andrews KY16 9SS Fife Scotland M Squared Lasers 1 Kelvin CampusWest Scotland Sci Pk Glasgow G20 0SP Lanark Scotland

An outstanding challenge for immunology is the classification of immune cells in a label-free fashion with high speed. For this purpose, optical techniques such as Raman spectroscopy or digital holographic microscopy have been used successfully to identify immune cell subsets. To achieve high accuracy, these techniques require a post-processing step using linear methods of multivariate processing, such as principal component analysis. Here we demonstrate for the first time a comparison between artificial neural networks and principal component analysis (PCA) to classify the key granulocyte cell lineages of neutrophils and eosinophils using both digital holographic microscopy and Raman spectroscopy. Artificial neural networks can offer advantages in terms of classification accuracy and speed over a PCA approach. We conclude that digital holographic microscopy with convolutional neural networks based analysis provides a route to a robust, stand-alone and high-throughput hemogram with a classification accuracy of 91.3 % at a throughput rate of greater than 100 cells per second. Published by The Optical Society under the terms of the Creative Commons Attributton 4.0 license.

关键词： Forward scattering Holographic microscopy image processing neural networks Raman scattering stochastic gradient descent

来源：评论

学校读者我要写书评

暂无评论

Wide context learning network for stereo matching

引用

signal processing-image COMMUNICATION 2019年 78卷 263-273页

作者： Tien Phuoc Nguyen Jeon, Jae Wook Sungkyunkwan Univ Dept Elect & Comp Engn Suwon South Korea

Binocular stereo matching is a challenging problem in computer vision. Recently, convolutional neural networks (CNNs) have emerged as a promising approach. However, matching ambiguities on ill-posed regions remain an intractable challenge for current methods. In this paper, we propose a novel network model consisting of two main parts: a wide context learning network and stacked encoder-decoder 2D CNNs with a spatial diffusion module. The first part leverages the power of a dilated convolutional layer and spatial pyramid pooling to extract global context information and constitute a matching cost volume. The second part performs contextual aggregation over this matching cost volume to optimize and smooth the matching cost. Finally, we estimate a disparity map by computing the probability of each disparity from the predicted matching cost. The proposed network model allows us to train end-to-end without any further post-processing or refinement. In the experiments, we evaluate our method using the synthesis Scene Flow and real-world KITTI datasets. Our proposed method achieves performance results that are competitive against state-of-the-art methods while maintaining a fast run time.

关键词： Stereo matching 3D reconstruction Matching cost Cost aggregation Convolutional neural network

来源：评论

学校读者我要写书评

暂无评论

Marathon Bib Number Recognition using Deep Learning 11

Marathon Bib Number Recognition using Deep Learning

引用

11th International Symposium on image and signal processing and Analysis (ISPA)

作者： Apap, Adrian Seychell, Dylan Univ Malta Dept Artificial Intelligence Msida Malta

ISBN: (纸本)9781728131405

Bib number recognition (BNR) from unstructured marathon images can be a challenging task. This is because the images captured at these events are very inconsistent since, they are often captured by multiple photographers, at various locations and times. This results in images containing different backgrounds, angles and illumination. The images often contain multiple participants in various poses, where the bib numbers can be obstructed by the participants themselves. The bib numbers are often printed on flexible paper and can easily be deformed which distorts the printed numbers. In this work we present a BNR system based on deep learning which is able to locate bib numbers in unstructured, complex marathon images. Using the segmented bib numbers the system then, recognizes the digits and finally outputs the bib numbers that it was able to detect in the image. The first stage consists of a fully Convolutional neural Network (CNN) to segment the bib numbers while the second stage consists of a Convolutional Recurrent neural Network (CRNN) used to recognize the detected numbers. The proposed method obtained an F1 score of 0.69 which outperformed existing methods.

关键词： Convolutional neural Network Segmentation Detection Recognition

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：