检索结果-内蒙古大学图书馆

IEEE Symposium on Nuclear Science (NSS/MIC)

作者： H. Li S. Yang P. Gan S. Mungai N. Ren Z. Kuang Y. Yang Z. Liu Paul C. Lauterbur Research Center for Biomedical Imaging Shenzhen Institute of Advanced Technology Shenzhen China

ISBN: (数字)9798350388152

ISBN: (纸本)9798350388169

Objective: This study aims to address the challenges associated with the traditional image reconstruction algorithms such as the numerous artifacts and poor spatial resolution for the simple back-projection (SBP) method, as well as the lengthy iteration times and extensive noise levels for the list mode maximum likelihood expectation maximization (LM-MLEM) method. A specialized reconstruction algorithm was developed and evaluated using data from a prototype high-precision Compton camera. methods: The Compton camera incorporates two layers of detectors. The scatter detector consists of an array of $21 \times 21$ GAGG crystals of $0.68 \times 0.68 \times 2 \mathrm{~mm}^{3}$ single-ended readout by an $8 \times 8$ SiPM array with an active pixel area of $2 \times 2 \mathrm{~mm}^{2}$. The absorber detector consists of an array of $23 \times 23$ LYSO crystals of $1.0 \times 1.0 \times 20 \mathrm{~mm}^{3}$ dual-ended readout by $8 \times 8$ SiPM arrays with an active pixel area of $3 \times 3 \mathrm{~mm}^{2}$. The proposed image reconstruction algorithm utilizes an inverse projection technique to incorporate prior image data and employs parallel computing along with a weight adjustment strategy. It is named the weight-adjusted SBP-based LM-MLEM algorithm (WA-SBP-MELM). This facilitates the computation and quantification of the weights of Compton projection rings at varying positions, thereby supporting the iterative calculations required for MLEM. Data acquired with the prototype Compton camera using a 0.3 mm diameter ${ }^{22} \mathrm{Na}$ point source will be used to compare different algorithms. Results: The proposed WA-SBP-MELM algorithm resulted in improvements in spatial resolution, noise performance, and reconstruction speed compared to previous algorithms. A spatial resolution of $\sim 0.9 \mathrm{~mm}$ was achieved for the point source. In the future, data of multiple point sources and phantoms will be acquired to further evaluate the proposed algorithm.

关键词： Semiconductor detectors Prototypes Phantoms Crystals Reconstruction algorithms parallel processing Cameras

来源：评论

学校读者我要写书评

暂无评论

Deep Learning for Medical image Segmentation 6th

Deep Learning for Medical Image Segmentation

引用

6th International conference on Business Intelligence, CBI 2021

作者： El Akraoui, Bouchra Daoui, Cherki Laboratory of Information Processing and Decision Support Faculty of Science and Technique Sultan Moulay Slimane University Beni-Mellal Morocco

ISBN: (纸本)9783030765071

Alzheimer’s disease is recognized as a progressive loss of memory and often leads to a total loss of autonomy, which renders it difficult to tolerate. Hippocampus is among the first cerebral regions to be infected in AD. An accurate diagnosis at an early stage of AD is crucial for the intervention process. The low contrast of supporting tissues and organs makes segmentation of the CC from brain MRI very difficult. In this paper we introduced both automatic brain image segmentation methods to extract the Corpus Callosum (CC), the first based on SLIC using parallel implementation which gives accelerated results over classical SLIC, the second method motivated by using Deep Learning approach. Finally, we compare the outcome given by the previous approaches. © 2021, Springer Nature Switzerland AG.

关键词： Magnetic resonance imaging

来源：评论

学校读者我要写书评

暂无评论

Region-of-Interest Based Neural Video Compression 33

Region-of-Interest Based Neural Video Compression

引用

33rd British Machine Vision conference Proceedings, BMVC 2022

作者： Perugachi-Diaz, Yura Sautière, Guillaume Abati, Davide Yang, Yang Habibian, Amirhossein Cohen, Taco Vrije Universiteit Amsterdam Netherlands Qualcomm AI Research Qualcomm Technologies Netherlands B.V. Netherlands Qualcomm AI Research Qualcomm Technologies Inc.

Humans do not perceive all parts of a scene with the same resolution, but rather focus on few regions of interest (ROIs). Traditional Object-Based codecs take advantage of this biological intuition, and are capable of non-uniform allocation of bits in favor of salient regions, at the expense of increased distortion the remaining areas: such a strategy allows a boost in perceptual quality under low rate constraints. Recently, several neural codecs have been introduced for video compression, yet they operate uniformly over all spatial locations, lacking the capability of ROI-based processing. In this paper, we introduce two models for ROI-based neural video coding. First, we propose an implicit model that is fed with a binary ROI mask and it is trained by de-emphasizing the distortion of the background. Secondly, we design an explicit latent scaling method, that allows control over the quantization binwidth for different spatial regions of latent variables, conditioned on the ROI mask. By extensive experiments, we show that our methods outperform all our baselines in terms of Rate-Distortion performance in the ROI. Moreover, they can generalize to different datasets and ROI specifications at inference time. Finally, they do not require expensive pixel-level annotations during training, as synthetic ROI masks can be used with little to no degradation in performance. To the best of our knowledge, our proposals are the first solutions that integrate ROI-based capabilities into neural video compression models. © 2022. The copyright of this document resides with its authors. It may be distributed unchanged freely in print or electronic forms.

关键词： image coding

来源：评论

学校读者我要写书评

暂无评论

Uplink D-MIMO processing Using Kalman Filter Combining

Uplink D-MIMO Processing Using Kalman Filter Combining

引用

GLOBECOM 2022 - 2022 IEEE Global Communications conference

作者： Ke Wang Helmersson Pål Frenger Anders Helmersson Ericsson Research Linköping Sweden Linköping University Linköping Sweden

ISBN: (纸本)9781665435413

We propose a decentralized subset method for optimal processing and combining of uplink signals in a distributed MIMO (D-MIMO) network. We further propose the use of Kalman filters with the square-root implementation to estimate the received uplink signals. This square-root implementation is shown to be numerically stable when inverting the covariance matrix, as it always assures the covariance matrix to be symmetric and positive semi-definite. In the paper we also analyze the computational complexity and cost with different combining methods. We show that the Kalman filter implementation provides the same result as the MMSE method in terms of the spectral efficiencies and equivalent SINR. However, the Kalman filter implementation is shown to be very efficient as it provides the possibility to fully utilize parallel computing of distributed hardware processors. Moreover, the processing can be decentralized and the estimates can be aggregated from local estimates to as many access points (APs) as needed to reach the desired performance target. A Kalman filter implementation has the flexibility to aggregate signals in different ways, allowing the fronthaul architecture to support connectivity of individual APs in any combination of parallel or serial manners.

关键词： Upper bound Costs Spectral efficiency Aggregates Stars Topology Kalman filters

来源：评论

学校读者我要写书评

暂无评论

Scalable Deep Learning on distributed Infrastructures: Challenges, Techniques, and Tools

引用

ACM COMPUTING SURVEYS 2020年第1期53卷 3-3页

作者： Mayer, Ruben Jacobsen, Hans-Arno Tech Univ Munich Boltzmannstr 3 D-85748 Garching Germany

Deep Learning (DL) has had an immense success in the recent past, leading to state-of-the-art results in various domains, such as image recognition and natural language processing. One of the reasons for this success is the increasing size of DL models and the proliferation of vast amounts of training data being available. To keep on improving the performance of DL, increasing the scalability of DL systems is necessary. In this survey, we perform a broad and thorough investigation on challenges, techniques and tools for scalable DL on distributed infrastructures. This incorporates infrastructures for DL, methods for parallel DL training, multi-tenant resource scheduling, and the management of training and model data. Further, we analyze and compare 11 current open-source DL frameworks and tools and investigate which of the techniques are commonly implemented in practice. Finally, we highlight future research trends in DL systems that deserve further research.

关键词： Deep-learning systems

来源：评论

学校读者我要写书评

暂无评论

DETECTION of FERTILE SOILS BASED on SATELLITE imageRY processing 9

DETECTION of FERTILE SOILS BASED on SATELLITE IMAGERY PROCES...

引用

9th International conference "distributed Computing and Grid Technologies in Science and Education", GRID 2021

作者： Grishkin, Valery Zhivulin, Evgeniy Khokhriakova, Anastasiia Karimov, Sardor Saint Petersburg State University 7-9 Universitetskaya nab. Saint Petersburg199034 Russia

The paper proposes a method for detecting fertile soils based on the processing of satellite images. As a result of its application, a map of the location of fertile and infertile soils for a given region of the earth's surface is formed and the corresponding areas are calculated. The method for detecting fertile soils is based on the fact that fertile soil includes areas covered with vegetation in the spring-summer period. Therefore, by measuring the spectral characteristics of these areas in the late autumn period, when there is no vegetation on them, it is possible to obtain objective parameters of fertile soils. For detection, a number of classifiers are being built that recognize two classes - fertile soil and sand, which is especially important when monitoring areas prone to desertification. The feature vector used for classification is a set of indices similar to the well-known NDVI index. This set of indices is calculated for each pixel of the image by its values in different spectral channels. Classifiers are implemented using CUDA parallel computing technology on a GPU. Based on the results of the experimental study, a classifier is selected that has shown the best characteristics of the recognition quality. Copyright © 2021 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).

关键词： image segmentation

来源：评论

学校读者我要写书评

暂无评论

Transformer in UAV image-Based Weed Mapping

引用

REMOTE SENSING 2023年第21期15卷 5165-5165页

作者： Zhao, Jiangsan Berge, Therese With Geipel, Jakob Norwegian Inst Bioecon Res NIBIO Dept Agr Technol POB 115 NO-1431 As Norway Norwegian Inst Bioecon Res NIBIO Dept Invertebrate Pests & Weeds Forestry Agr & Hort POB 115 NO-1431 As Norway

Weeds affect crop yield and quality due to competition for resources. In order to reduce the risk of yield losses due to weeds, herbicides or non-chemical measures are applied. Weeds, especially creeping perennial species, are generally distributed in patches within arable fields. Hence, instead of applying control measures uniformly, precision weeding or site-specific weed management (SSWM) is highly recommended. Unmanned aerial vehicle (UAV) imaging is known for wide area coverage and flexible operation frequency, making it a potential solution to generate weed maps at a reasonable cost. Efficient weed mapping algorithms need to be developed together with UAV imagery to facilitate SSWM. Different machine learning (ML) approaches have been developed for image-based weed mapping, either classical ML models or the more up-to-date deep learning (DL) models taking full advantage of parallel computation on a GPU (graphics processing unit). Attention-based transformer DL models, which have seen a recent boom, are expected to overtake classical convolutional neural network (CNN) DL models. This inspired us to develop a transformer DL model for segmenting weeds, cereal crops, and 'other' in low-resolution RGB UAV imagery (about 33 mm ground sampling distance, g.s.d.) captured after the cereal crop had turned yellow. images were acquired during three years in 15 fields with three cereal species (Triticum aestivum, Hordeum vulgare, and Avena sativa) and various weed flora dominated by creeping perennials (mainly Cirsium arvense and Elymus repens). The performance of our transformer model, 1Dtransformer, was evaluated through comparison with a classical DL model, 1DCNN, and two classical ML methods, i.e., random forest (RF) and k-nearest neighbor (KNN). The transformer model showed the best performance with an overall accuracy of 98.694% on pixels set aside for validation. It also agreed best and relatively well with ground reference data on total weed coverage, R2 = 0.598. I

关键词： 1Dtransformer perennial weed mapping low-resolution UAV image

来源：评论

学校读者我要写书评

暂无评论

Building Deep Neural Networks for solving Machine Learning Problems in Agricultural Production

Building Deep Neural Networks for solving Machine Learning P...

引用

International Russian Smart Industry conference (SmartIndustryCon)

作者： Aleksey Rogachev Elena Melikhova Natalia Zolotykh Department of Mathematical Modeling and Informatics Volgograd State Agricultural University Volgograd Russian Federation

In the tasks of agricultural production, it is necessary to identify unfavorable situations of agricultural farming that arise in the process of cultivating agricultural crops. These include soil erosion or salinization, damage from crop diseases, pests, and others. Timely and prompt identification of such situations is possible with the use of technical vision and methods of intellectual analysis and image processing. The most effective means of machine learning (ML) for such tasks are deep neural networks (DNN), primarily based on a parallel architecture containing convolutional layers of neurons. The purpose of the study was to build and study the effectiveness of DNN, which are used in intellectual land use tasks. The Python-based Google Collaboration cloud service, including ML libraries, was used as the DNN development environment.. When designing DNN, the features of the functioning of the CPU and GPU were taken into account. The results obtained make it possible to optimize the architecture and hyperparameters of DNN, as well as their training time. This approach increases the efficiency of the information and analytical complexes being developed to support the solution of various land use problems.

关键词： Training Deep learning Convolutional codes Neurons Graphics processing units Crops Production

来源：评论

学校读者我要写书评

暂无评论

Research on parallel k-nearest neighbor query optimization method for highway big data 2

Research on parallel k-nearest neighbor query optimization m...

引用

2nd IEEE Asia-Pacific conference on image processing, Electronics and Computers, IPEC 2021

作者： Li, Min Jingchu University of Technology School of Computer Engineering Jingmen Hubei448000 China

ISBN: (纸本)9781728190181

Because the information in the field of engineering construction involves remote sensing recognition, intelligent transportation, etc. The rapid growth of various types of information makes highway data continue to develop toward the trend of big data in terms of scale, type, and structure, showing the characteristics of multi-source heterogeneous. This article discusses efficient data retrieval and query methods for specific highway basic databases and simulated highway spatial large data sets. In order to better manage spatial objects and maintain their positional relationship with grid cells, it is proposed to combine inverted index with grid index. In order to break the dependence of the rim radius on the previous traversal results, the serial rim access K nearest neighbor (KNN) algorithm is appropriately improved, and the rim radius is updated by increasing the mesh side length, combined with multi-threading technology Achieve access to multiple rims to improve the query performance of highway big data. Test data by modifying the size, spatial grid sides and these K values, the average performance are improved. © 2021 IEEE.

关键词： Big data

来源：评论

学校读者我要写书评

暂无评论

A distributed Cache Mechanism of HDFS to Improve Learning Performance for Deep Reinforcement Learning

A Distributed Cache Mechanism of HDFS to Improve Learning Pe...

引用

IEEE International conference on Big Data and Cloud Computing (BdCloud)

作者： Yongqiang Gao Shunyi Feng Zhenkun Li College of Computer Science Inner Mongolia University Hohhot China

Hadoop distributed file system (HDFS) has been widely used in deep reinforcement learning due to its high reliability and high performance. However, the dataset for training deep reinforcement learning models is usually composed of a large number of small files, which will cause the severe performance degradation of HDFS. Although many methods have been proposed to address small files problem in HDFS, none of them considers correlations between small files for training deep reinforcement learning models. To improve the learning performance for deep reinforcement learning, this paper proposes a new HDFS architecture, which uses a distributed cache mechanism to improve data access efficiency for training deep reinforcement learning models. The experimental results show that the proposed HDFS with distributed cache mechanism can achieve better learning performance compared to the traditional HDFS without cache and with centralized caching mechanism.

关键词： Deep learning Training Degradation Energy consumption File systems Clustering algorithms distributed databases

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：