检索结果-内蒙古大学图书馆

IEEE Symposium on Nuclear Science (NSS/MIC)

作者： E. Cece P. v.F.J. Meyrat O. verdier E. Torino M. Colarieti-Tosti Division of Biomedical Imaging KTH Royal Institute of Technology Stockholm Sweden Department of Chemical Engineering Materials and Production University of Naples Federico II Naples Italy Department of Computing Mathematics and Physics HVL Western Norway University of Applied Sciences Bergen Norway Centro Interdipartimentale di Ricerca sui Biomateriali (CRIB) University of Naples Federico II Naples Italy

The diagnostic power of Positron Emission Tomography for early stage lung cancer is challenging since the lesion dimensions are comparable with the spatial resolution of modern clinical scanners and their contrast is decreased by respiratory motion. Moreover, strong artifacts in reconstructed images are introduced by attenuation, resulting in depression of activity in the middle of the image. In this paper we propose a modified version of Maximum Likelihood Activity and Attenuation (MLAA) algorithm, the Morphed-MLAA (M-MLAA), which, given gated data, jointly corrects for patient motion during the acquisition by composing the forward and backward operators with the registration fields between the different gates. The registration fields are evaluated with the deep-learning model Synthmorph. The main advantage of the proposed algorithm is that it makes use of the entire projection data set, thus improving the lesion signal-to-noise ratio, compared to the case in which images are reconstructed from a subset of the data. Our proposed algorithm has been tested on the XCAT phantom with some added lung lesions and shows improved performance for the detection and reconstruction of some of them; this improved performance is corroborated by the evaluation of two figures of merit (PSNR and RC). Our results show the potential of M-MLAA algorithm, but, at the same time, also that there is still space for improvements, since some lesions are better detected by the other more classical approaches, as MLAA single or total gates reconstructions.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Experimental Evaluation of IoT based Human Gender Classification and Record Management using Intelligent Hybrid Learning Principles

Experimental Evaluation of IoT based Human Gender Classifica...

引用

Advances in Computational Intelligence and Communication (ICACIC), International conference on

作者： P vinayagam v Nithesh Kumar G Bhuvan Department of ECE Saveetha Engineering College Chennai India

There has been a lot of recent interest in automatically predicting age and gender from facial images because to the several uses this has in different kinds of facial investigations. Using the aforementioned technology, we can tell a person’s age and gender from only one image. In this research, we develop a unique deep learning approach, Intelligent Hybrid Learning Principles (IHLP), and apply it to the task of classifying facial images according to gender. Convolutional Neural Networks (CNNs), a standard learning technique, serve as the inspiration for the suggested logic. Both models were tested in experiments designed to see how well they could distinguish between male and female faces. To approximately estimate a human face’s gender and age from a picture, the primary goal of this research is to apply a deep learning based IHLP to construct a gender and age detector. In order to accomplish a gender classification/identification job, automatic face recognition seeks to extract the important bits of information from humans and combine them into a usable representation. All of them, however, have something holding them back, such as inaccurate or incomplete reflection of the face’s structure or texture. Both alignment of faces and identification of faces may benefit greatly from this method. We propose a straightforward convolutional neural network design based on Intelligent Hybrid Learning Principles for this purpose, which can be used even in the face of a scarcity of training data. The effectiveness of the suggested method is shown by cross-validation using a model of a convolutional neural network, which then serves as evidence for the resultant section. Apart from this an advanced logic of Internet of Things (IoT) is adopted over this approach to maintain the records clearly for further analysis and it is easy to recover the records for further use.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Unsupervised Singing voice Detection Using Dictionary Learning

Unsupervised Singing Voice Detection Using Dictionary Learni...

引用

European Signal Processing conference

作者： Aggelos Pikrakis Yannis Kopsinis Nadine Kroher Jose-Miguel Diaz-Banez Dept. of Informatics University of Piraeus Greece Libra MLI Edinburgh UK Dept. of Applied Mathematics II University of Seville Spain

ISBN: (纸本)9781509018918

This paper presents an unsupervised approach to vocal detection in music recordings based on dictionary learning. At a first stage, the recording to be segmented is treated as training data and the K-SvD algorithm is used to learn a dictionary which sparsely represents a short-term feature sequence that has been extracted from the recording. Subsequently, the vectors of the feature sequence are reconstructed based on the learned dictionary and the probability of appearance of the dictionary atoms is estimated. The obtained probability serves to compute the value of a weight function for each frame of the recording. The histogram of this function is then used to estimate a binarization threshold that segments the recording into vocal and non-vocal segments. The performance of the proposed unsupervised method, when evaluated on two datasets of accompanied singing, presents comparable performance to supervised techniques.

关键词： Dictionaries Histograms Feature extraction image reconstruction Training Signal processing Europe

来源：评论

学校读者我要写书评

暂无评论

Transfer-Learning and YOLO v7 Hybridised for Human Cropping for Deepfake Detection Algorithms

Transfer-Learning and YOLO V7 Hybridised for Human Cropping ...

引用

IEEE International conference on Smart Instrumentation, Measurement and Applications (ICSIMA)

作者： Nawaf Waqas Sairul Izwan Safie Kushsairy A. Kadir Sheroz Khan Abdul Ali Khan Universiti Kuala Lumpur British Malaysian Institute Kuala Lumpur Malaysia Universiti Kuala Lumpur Malaysian Institute of Industrial Technology Johor Malaysia College of Engineering and Information Technology Onaizah Colleges Saudi Arabia

The advent of deep learning technology has created images and videos convincingly fake to appeal viewers enough in appearing close to real. Such image and video modalities, known as Deepfakes, can cause harm to individuals, communities, and countries. Detecting Deepfakes is essential to prevent their negative impacts, and several detection techniques that utilize deep learning algorithms have been developed over the recent past years. The algorithms reduce the target modality to dimensional latent space to include features important for subsequent reconstruction of the same modality from the developed latent distribution data. However, manual data preprocessing such as splitting frames and cropping regions from Deepfake videos for the facial features can be tedious and time-consuming. This article proposes YOLO v7 hybridised with transfer-learning to automate the cropping process in Deepfake videos and frames splitting, making it easier to extract facial features from videos. various popular datasets have been used for Deepfake detection, and this article explores the benefits of the proposed approach. The results present a comparison of the proposed approach with state-of-the-art models from the levels of Precision, Recall, and mAP@.5 to demonstrate the robustness of the proposed technique. Notably, it is evident that our approach outperforms other models, boasting a precision value of 0.9009, a recall value of 0.8979, and an mAP@.5 value of 0.8745.

关键词：

来源：评论

学校读者我要写书评

暂无评论

virtual exhibitions as a collection environment of cultural events

Virtual exhibitions as a collection environment of cultural ...

引用

IEEE International conference on Multimedia Computing and Systems (ICMCS)

作者： G. Garibotto M.v. Oneto C. Coelho M. Marson G. Tartaglione L. Maccabruni Elsag S.p.A. Genoa Italy Archivio di Stato France

Multimedia technology provides improved solutions to allow access to distributed resources and data in a variety of applications. There are examples of implementation in different domains, from commercial distribution through electronic commerce to virtual museums and dissemination of cultural information. In particular in the Cultural Heritage area there are already many different implementations of high quality services which are distributed either through specialised CD's or through the Internet. The paper refers an experience which has been performed by Elsag SpA, in cooperation with Florence Archivio di State, to exploit the use of multimedia technology in the area of virtual Exhibition, in order to maintain a memory of temporary events which otherwise, after the closure of the event itself are bound to disappear. The experiment has been implemented with the use of consolidated tools and techniques and demonstrates the maturity of available technology as well as its limitations and need of improvement.

关键词： Cultural differences Layout image reconstruction Space technology Workstations Electronic commerce Web and internet services Paper technology virtual environment Time factors

来源：评论

学校读者我要写书评

暂无评论

sLLE: Spherical Locally Linear Embedding with Applications to Tomography

sLLE: Spherical Locally Linear Embedding with Applications t...

引用

IEEE conference on Computer vision and Pattern Recognition

作者： Yi Fang Mengtian Sun S. v. N. vishwanathan Karthik Ramani Purdue University

ISBN: (纸本)9781457703942

The tomographic reconstruction of a planar object from its projections taken at random unknown view angles is a problem that occurs often in medical imaging. Therefore, there is a need to robustly estimate the view angles given random observations of the projections. The widely used locally linear embedding (LLE) technique provides nonlinear embedding of points on aflat manifold. In our case, the projections belong to a sphere. Therefore, we extend LLE and develop a spherical locally linear embedding (sLLE) algorithm, which is capable of embedding data points on a non-flat spherically constrained manifold. Our algorithm, sLLE, transforms the problem of the angle estimation to a spherically constrained embedding problem. It considers each projection as a high dimensional vector with dimensionality equal to the number of sampling points on the projection. The projections are then embedded onto a sphere, which parametrizes the projections with respect to view angles in a globally consistent manner. The image is reconstructed from parametrized projections through the inverse Radon transform. A number of experiments demonstrate that sLLE is particularly effective for the tomography application we consider. We evaluate its performance in terms of the computational efficiency and noise tolerance, and show that sLLE can be used to shed light on the other constrained applications of LLE.

关键词： look angles (electronics) embedding Tomography PROJECTION noise tolerance Manifolds Spheres Computational efficiency

来源：评论

学校读者我要写书评

暂无评论

Codebook vQ-vAE Approach for Prostate Cancer Diagnosis using Multiparametric MRI

Codebook VQ-VAE Approach for Prostate Cancer Diagnosis using...

引用

IEEE Computer Society conference on Computer vision and Pattern Recognition Workshops (CvPRW)

作者： Ekaterina Redekop Mara Pleasure Zichen Wang Karthik v Sarma Adam Kinnaird William Speier Corey W Arnold University of California Los Angeles USA University of Alberta Canada

ISBN: (数字)9798350365474

ISBN: (纸本)9798350365481

Multiparametric magnetic resonance imaging (mpMRI) plays an essential role in prostate cancer diagnosis as it can noninvasively localize and grade lesions based on their suspicion of representing clinically significant prostate cancer (csPCa). With the development of deep learning, automatic solutions for csPCa detection based on mpMRI have been developed; however, mpMRI data introduces several difficulties, including data scarcity, heterogeneity in image quality across institutions, and missing modalities. This work addresses these difficulties by building a radiology-based foundational model for prostate cancer mpMRI. Foundation models are deep learning models pretrained on a large-scale dataset and they have recently gained significant interest in computer vision and natural language applications. After pretraining, these models are often adapted for a variety of downstream tasks using smaller datasets from within the same domain. In this work, a large prostate multiparametric MRI (mpMRI) dataset was collected by combining data from our institution with two publicly available datasets. Joint modeling of all mpMRI modalities is essential for accurate prostate cancer diagnosis; however, some of these modalities may be missing. Using unsupervised learning, we pretrained modality-specific vector quantized variational autoencoders (vQ-vAE) to form a radiology foundational model. The learned codebook from vQ-vAE was then used to train a multimodal transformer to perform the diagnosis of clinically significant prostate cancer (csPCa). The proposed multimodal transformer models long-range dependencies between latent representations of input modalities and is augmented with modality-level dropout to increase the model robustness to incomplete modalities. Our framework outperforms previously published work and achieves an average AUC/sensitivity/specificity of 0.764/0.690/0.781. Our results show that pretraining on a larger dataset in combination with the power of transfor

关键词： Deep learning Adaptation models Computer vision Frequency modulation Computational modeling Magnetic resonance imaging Radiology

来源：评论

学校读者我要写书评

暂无评论

Comparison of genetic-algorithm and emissivity-ratio analyses of image data from OMEGA implosion cores

引用

REvIEW OF SCIENTIFIC INSTRUMENTS 2008年第10期79卷 10E921页

作者： Nagayama, T. Mancini, R. C. Florido, R. Tommasini, R. Koch, J. A. Delettrez, J. A. Regan, S. P. Smalyuk, v. A. Welser-Sherrill, L. A. Golovkin, I. E. Univ Nevada Dept Phys Reno NV 89557 USA Lawrence Livermore Natl Lab Livermore CA 94550 USA Univ Rochester Laser Energet Lab Rochester NY 14623 USA Los Alamos Natl Lab Los Alamos NM 87545 USA Prism Computat Sci Madison WI 53711 USA

Detailed analysis of x-ray narrow-band images from argon-doped deuterium-filled inertial confinement fusion implosion experiments yields information about the temperature spatial structure in the core at the collapse of the implosion. We discuss the analysis of direct-drive implosion experiments at OMEGA, in which multiple narrow-band images were recorded with a multimonochromatic x-ray imaging instrument. The temperature spatial structure is investigated by using the sensitivity of the Ly beta/He beta line emissivity ratio to the temperature. Three analysis methods that consider the argon He beta and Ly beta image data are discussed and the results compared. The methods are based on a ratio of image intensities, ratio of Abel-inverted emissivities, and a search and reconstruction technique driven by a Pareto genetic algorithm.

关键词： explosions genetic algorithms laser fusion plasma diagnostics plasma temperature

来源：评论

学校读者我要写书评

暂无评论

Small-sized computed tomography system with rotating gantry

Small-sized computed tomography system with rotating gantry

引用

International conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology, ECTI-CON

作者： v. Jantanayingyong T. Chanwimalueang N. Srisuk C. Pintavirooj Faculty of Engineering King Mongkut's Institute of Technology Ladkrabang (KMITL) Thailand

ISBN: (纸本)9781424456062;9789746724913

Small-sized computed tomography system with rotating gantry is presented in this paper. System calibration was performed before achieving the projection data. An object placed on platform was exposed by the x-ray every two degrees of rotating angle. Thus, 180 projection data were performed in this study. Finally, Feldkamp algorithm based on a 3D filtered backprojeciton was employed to reconstruct cross sectional images. The result shows that cross sectional images obtained from small-sized computed tomography were satisfactory. In conclusion, the small-sized computed tomography presented in this study is capable to construct the cross sectional images. Moreover, 3D model reconstructed from these cross sectional images will be performed in the future work.

关键词： Computed tomography X-ray imaging image reconstruction Geometry X-ray detectors X-ray detection Calibration Costs Medical services Biomedical equipment

来源：评论

学校读者我要写书评

暂无评论

Qualitative model of spatio-temporal radiation pattern of GPR antenna

Qualitative model of spatio-temporal radiation pattern of GP...

引用

Proceedings of the International conference on Ground Penetrating Radar, GPR

作者： F. D. Edemskiy A. v. Popov S. A. Zapunidi Department of radio wave propagation Ionosphere and Radio Wave Propagation Pushkov Institute of Terrestrial Magnetism Troitsk Moscow Russia

In our recent work, a simple qualitative model of a dipole GPR antenna radiation pattern has been developed. Replacing the real resistively-loaded dipole by an infinite line current source stretched along the ground-air interface reduces derivation to a 2D boundary value problem for a scalar wave equation which admits an exact analytical solution. Using the complex variable technique we express the Green function in terms of elementary functions and obtain the spatio-temporal radiation pattern for an arbitrary current pulse form by Duhamel integral. In spite of this simplification, the constructed solution describes quite well the ultra-wideband (UWB) pulse emitted by a realistic GPR antenna. These analytical results are useful for the interpretation of GPR raw data by giving a full description of the pulse amplitude and waveform dependence on the radiation angle. Moreover, an inverse problem can be solved: to find the unknown antenna current pulse from the signal measured in the air half-space. Two cases are considered: the surface wave and the aerial wave propagating upwards. This information gives the possibility to deconvolve the subsurface probing data thus increasing the GPR resolution.

关键词： Ground penetrating radar Antenna radiation patterns Antenna measurements Green function image reconstruction Dipole antennas

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：