检索结果-内蒙古大学图书馆

7th Optics Young Scientist Summit, OYSS 2024

作者： Sun, Bingheng Ren, Zhong Liang, Gaoqiang Jiangxi Provincial Key Laboratory of Advanced Electronic Materials and Devices Jiangxi Science and Technology Normal University Nanchang330038 China Key Laboratory of Optic-electronic Detection and Information Processing of Nanchang City Jiangxi Science and Technology Normal University Nanchang330038 China

ISBN: (数字)9781510688377

ISBN: (纸本)9781510688360

Photoacoustic imaging is a kind of rapidly advancing noninvasive detection technology based on the photon-induced-acoustic effect, which holds the significant potential applications in the biomedical imaging diagnosis field. This technique merges the high contrast capabilities of optical imaging with the deep penetration and high resolution of acoustic imaging. In our work, to study the interaction mechanisms and photoacoustic effects between pulsed laser and capillaries in skin tissue, finite element analysis (FEA) based on COMSOL multiphysis software was employed. Based on the established finite element bio-tissue model, the photoacoustic effect was fully and completely represented as a sequence of light absorption, thermal expansion, and ultrasound wave generation in the bio-tissue, involving the coupling of multiple physical fields and the numerical resolution of partial differential equations. Moreover, in our study, the effects of varying induced wavelengths, capillaries depths in tissue, and absorber sizes on the time-domain photoacoustic signals and photoacoustic intensity were numerical investigated. Numerical results show that the blood in capillaries has stronger photoacoustic effect in the visible light waveband due to the characteristic absorption of hemoglobin in capillaries. The photoacoustic intensity decreases with the increase of capillary radius and depth. To validate the availability of the numerical simulation, the corresponding photoacoustic experiments were performed. In the experiments, we constructed a photoacoustic microscopy detection system and experimentally studied the photoacoustic effects of several influence factors. The experimental results confirm the accuracy of our numerical simulations and provide theoretical insights for future applications of vascular or capillaries photoacoustic imaging in the biomedical field. © 2025 SPIE.

关键词： Tissue

来源：评论

学校读者我要写书评

暂无评论

An overview on IRS-enabled sensing and communications for 6G: architectures, fundamental limits, and joint beamforming designs

引用

Science China(information Sciences) 2025年第5期68卷 170-193页

作者： Xianxin SONG Yuan FANG Feng WANG Zixiang REN Xianghao YU Ye ZHANG Fan LIU Jie XU Derrick Wing Kwan NG Rui ZHANG Shuguang CUI School of Science and Engineering (SSE) Shenzhen Future Network of Intelligence Institute (FNii-Shenzhen)and Guangdong Provincial Key Laboratory of Future Networks of Intelligence The Chinese University of Hong Kong Department of Electrical Engineering City University of Hong Kong School of Information Engineering Guangdong University of Technology Key Laboratory of Wireless-Optical Communications Chinese Academy of Sciences School of Information Science and Technology University of Science and Technology of China Computer School Beijing Information Science and Technology University National Mobile Communications Research Laboratory School of Information Science and Engineering Southeast University School of Electrical Engineering and Telecommunications University of New South Wales School of Science and Engineering Shenzhen Research Institute of Big Data The Chinese University of Hong Kong Department of Electrical and Computer Engineering National University of Singapore

This study presents an overview on intelligent reflecting surface(IRS)-enabled sensing and communication for the forthcoming sixth-generation(6G) wireless networks, in which IRSs are strategically deployed to proactively reconfigure wireless environments to improve both sensing and communication(S&C) performance. First, we exploit a single IRS to enable wireless sensing in the base station's(BS's) non-line-of-sight(NLoS) area. In particular, we present three IRS-enabled NLoS target sensing architectures with fully-passive, semi-passive, and active IRSs, respectively. We compare their pros and cons by analyzing the fundamental sensing performance limits for target detection and parameter estimation. Next, we consider a single IRS to facilitate integrated sensing and communication(ISAC), in which the transmit signals at the BS are used for achieving both S&C functionalities, aided by the IRS through reflective beamforming. We present joint transmit signal and receiver processing designs for realizing efficient ISAC, and jointly optimize the transmit beamforming at the BS and reflective beamforming at the IRS to balance the fundamental performance tradeoff between S&C. Furthermore, we discuss multi-IRS networked ISAC, by particularly focusing on multi-IRS-enabled multi-link ISAC, multi-region ISAC, and ISAC signal routing, respectively. Finally, we highlight various promising research topics in this area to motivate future work.

关键词： integrated sensing and communication (ISAC) intelligent reflecting surface (IRS) non-line-of-sight (NLoS) sensing sensing and communication tradeoff

来源：评论

学校读者我要写书评

暂无评论

CFinDEE: A Chinese Fine-Grained Financial Dataset for Document-Level Event Extraction 24

CFinDEE: A Chinese Fine-Grained Financial Dataset for Docume...

引用

33rd Companion of the ACM World Wide Web Conference, WWW 2023

作者： Zhang, Tian Liu, Maofu Zhou, Bingying School of Computer Science and Technology Hubei Province Key Laboratory of Intelligent Information Processing and Real-time Industrial System Wuhan University of Science and Technology Wuhan China

ISBN: (纸本)9798400701726

Document-level event extraction faces numerous challenges in accurately modeling real-world financial scenarios, particularly due to the inadequacies in existing datasets regarding data scale and fine-grained annotations. The development of datasets is a crucial factor in driving research progress;therefore, we present a high-quality Chinese document-level event extraction dataset, CFinDEE. This dataset, grounded in real-world financial news, defines 22 event types and 116 argument roles, annotating 26,483 events and 107,096 event arguments. CFinDEE aims to address these shortcomings by providing more comprehensive annotations and data augmentation, offering richer resources for document-level event extraction in the financial domain. CFinDEE extends data both horizontally and vertically, where horizontal expansion enriches the types of financial events, enhancing the diversity of the dataset;vertical expansion, by increasing the scale of the data, effectively boosts the practical value of the dataset. Experiments conducted on multiple advanced models have validated the high applicability and effectiveness of the CFinDEE dataset for document-level event extraction tasks in the financial field. © 2024 Copyright held by the owner/author(s). Publication rights licensed to ACM.

关键词： Expansion

来源：评论

学校读者我要写书评

暂无评论

The Methodology of Mask Detection Employing the YOLOv7 Framework 4

The Methodology of Mask Detection Employing the YOLOv7 Frame...

引用

4th International Conference on Neural Networks, information and Communication Engineering, NNICE 2024

作者： Zhang, Fucheng Tian, Pinfang Wuhan University of Science and Technology School of Computer Science and Technology Wuhan China Hubei Key Laboratory of Intelligent Information Processing and Real-TimeIndustrial Systems Wuhan China

ISBN: (纸本)9798350394375

In the current field of object detection, the YOLOv7 model demonstrates significant advantages in terms of both detection speed and accuracy. However, the model shows deficiencies in focussing on key information, and its loss function's sensitivity to bounding boxes results in suboptimal performance in densely populated environments and small target detection scenarios. To address these issues, this study proposes a mask target detection method based on YOLOv7. By integrating a channel attention mechanism into the YOLOv7 model, this method enhances the model's focus on important information in images, effectively reducing interference from background factors. Additionally, the method incorporates Normalized Wasserstein Distance (NWD) into the loss function to optimize the model's evaluation of small target detection. Experimental verification shows that the improved model maintains its original detection speed while increasing the overall accuracy to 92.1%, a 2.9% improvement over the original YOLOv7 model. It achieves a 3.3% increase in the mAP.5 metric. This indicates that the optimised YOLOv7 model has superior detection capabilities, allowing it to perform mask-wearing detection tasks more rapidly and accurately in natural environments. © 2024 IEEE.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Dilation-Supervised Learning: A Novel Strategy for Scale Difference in Retinal Vessel Segmentation

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年第4期5卷 1693-1707页

作者： Wang, Huadeng Zuo, Wenbin Li, Bingbing Pan, Xipeng Liu, Zhenbing Lan, Rushi Luo, Xiaonan Guangxi Key Laboratory of Image and Graphic Intelligent Processing Guilin541004 China School of Computer Science and Information Security Guilin University of Electronic Technology Guilin541004 China Guilin University of Electronic Technology School of Computer Science and Information Security Guilin541004 China Ganzhou Municipal Hospital Department of Pathology Ganzhou341000 China

Retinal fundus image segmentation based on deep learning is an important method for auxiliary diagnosis of ophthalmic diseases. Due to the scale difference of the blood vessels and the imbalance between foreground and background pixels in the fundus image, the deep learning network will inevitably ignore thin vessels when downsampling and feature learning. For the scale difference problem, this article aims to tackle its limitation from two perspectives: changing the supervised approach and adapting the feature learning. Correspondingly, a dilation-supervised learning method and an adaptive scale dimensional attention mechanism which are used to construct a two-stage segmentation model is proposed. Moreover, we introduce a quantitative approach to evaluate the scale difference of the blood vessels. With the help of the proposed weighted loss function, the segmentation results are refined, and the class imbalance problem between foreground and background pixels is resolved. Finally, the proposed adaptive threshold selection method is used in the postprocessing of segmentation results. The experiments on DRIVE, STARE, CHASE_DB1, and HRF datasets show that the proposed method achieves better segmentation performance compared with other state-of-the-art methods, and has good generalization ability and robustness. © 2020 IEEE.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

UAV Image Matching Based on Graph Neural Network 16th

UAV Image Matching Based on Graph Neural Network

引用

16th International Conference on Genetic and Evolutionary Computing, ICGEC 2024

作者： Qiqi, Li Zhuo, Liu Lingyue, Meng Xiaomin, Liu Huaqi, Zhao The Heilongjiang Provincial Key Laboratory of Autonomous Intelligence and Information Processing School of Information and Electronic Technology Jiamusi University Heilongjiang Jiamusi154000 China College of Information and Electrical Engineering Heilongjiang Bayi Agricultural University Heilongjiang Daging163000 China

ISBN: (纸本)9789819615308

Unmanned aerial vehicle(UAV) image matching methods have been extensively applied in various fields such as search and rescue, precision agriculture, etc. Due to the change of the UAV attitude and complexity of flight environment, there are scale morphological changes and multi-view differences between the UAV images and satellite images. This paper proposed a UAV image matching based on graph neural network. First, a method of feature extraction based on multi-type attention graph neural network is introduced to solve the view differences between two images. Then, a multiple kernel clustering based on consensus affinity graph is used to cluster the feature descriptors, which deals with the problem of scale morphological changes. Finally, an optimal transport matching method based on random sampling is proposed to improve the robustness to noise. The experiment results demonstrate that our method outperforms the available methods. The AR of it is 1.97% higher than the original matching model. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Image matching

来源：评论

学校读者我要写书评

暂无评论

Lithological Facies Classification Using Attention-Based Gated Recurrent Unit

引用

Tsinghua Science and technology 2024年第4期29卷 1206-1218页

作者： Yuwen Liu Yulan Zhang Xingyuan Mao Xucheng Zhou Jingwen Chang Wenwei Wang Pan Wang Lianyong Qi College of Computer Science and Technology China University of Petroleum(East China)Qingdao 266580China Shandong Provincial University Laboratory for Protected Horticulture Weifang University of Science and TechnologyShouguang 262700China School of Mechanical Electrical and Information EngineeringShandong UniversityWeihai 264209China Shandong Provincial Key Laboratory of Network Based Intelligent Computing University of JinanJinan 250022China School of Modern Posts Nanjing University of Posts and TelecommunicationsNanjing 210023China

Lithological facies classification is a pivotal task in petroleum geology, underpinning reservoir characterization and influencing decision-making in exploration and production operations. Traditional classification methods, such as support vector machines and Gaussian process classifiers, often struggle with the complexity and nonlinearity of geological data, leading to suboptimal performance. Moreover, numerous prevalent approaches fail to adequately consider the inherent dependencies in the sequence of measurements from adjacent depths in a well. A novel approach leveraging an attention-based gated recurrent unit (AGRU) model is introduced in this paper to address these challenges. The AGRU model excels by exploiting the sequential nature of well-log data and capturing long-range dependencies through an attention mechanism. This model enables a flexible and context-dependent weighting of different parts of the sequence, enhancing the discernment of key features for classification. The proposed method was validated on two publicly available datasets. Results demonstrate a considerably improvement over traditional methods. Specifically, the AGRU model achieved superior performance metrics considering precision, recall, and F1-score.

关键词： attention mechanism facies classification GRU MLP(multilayer perceptron)

来源：评论

学校读者我要写书评

暂无评论

Intelligent Test Paper Generation Algorithm Integrating Normal Distribution and Parallel Genetic Algorithm 5

Intelligent Test Paper Generation Algorithm Integrating Norm...

引用

5th International Conference on Intelligent Computing and Human-computer Interaction, ICHCI 2024

作者： Liu, Chuangye Huang, Guimin School of Computer Science and Information Security Guilin University of Electronic Technology Guangxi Key Laboratory of Image and Graphic Intelligent Processing Guilin541004 China

ISBN: (纸本)9798350368284

Genetic algorithms have been widely used in intelligent test paper generation systems. However, traditional genetic algorithms cannot ensure that the difficulty of test questions is normally distributed, and are prone to falling into local optimal solutions. To address the above problems, we proposed an intelligent test paper generation algorithm that combines normal distribution and parallel genetic algorithms. In the population initialization stage, we use the cumulative distribution function (CDF) of the normal distribution to initialize the population. In the subsequent genetic operations, we designed a directed mutation mechanism and added KL divergence as a penalty number in the design of the fitness function to ensure that the difficulty of the test questions is normally distributed. At the same time, in order to speed up the efficiency of the algorithm and escape from the local optimal solution, an adaptive migration strategy was introduced. Experimental results show that the difficulty of the test questions generated by this algorithm is normally distributed, and the algorithm performance is better than other algorithms. © 2024 IEEE.

关键词： Consensus algorithm

来源：评论

学校读者我要写书评

暂无评论

引用

4th International Conference on computer Science and Blockchain, CCSB 2024

作者： Huang, Guimin Shi, Canjun School of Computer Science and Information Security Guilin University of Electronic Technology Guangxi Key Laboratory of Image and Graphic Intelligent Processing Guilin541004 China

ISBN: (纸本)9798350356502

Short textual similarity has been more and more widely used in the field of natural language processing. However, people tend to pay more attention to the semantic similarity of sentences and ignore the word order similarity in sentences. For this reason, we propose a short text similarity computation model that fuses semantics and word order. In terms of semantics, we propose a adjacency-aware similarity model based on deep convolutional neural networks, which effectively exploits the adjacency relationship in sentences and extracts text features from multiple perspectives for similarity prediction. In terms of word order, a word order model incorporating an external knowledge base is proposed, which uses an external knowledge base with pointwise mutual information to construct word order dense vectors to assess word order similarity. Ultimately, we propose a comprehensive assessment model for short text similarity that incorporates semantic and word order information. We conducted experiments on the public dataset MRPC, and the experimental results show that our model outperforms existing models in terms of performance. © 2024 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Inference-Time Decontamination: Reusing Leaked Benchmarks for Large Language Model Evaluation

Inference-Time Decontamination: Reusing Leaked Benchmarks fo...

引用

2024 Conference on Empirical Methods in Natural Language processing, EMNLP 2024

作者： Zhu, Qin Cheng, Qinyuan Peng, Runyu Li, Xiaonan Liu, Tengxiao Peng, Ru Qiu, Xipeng Huang, Xuanjing School of Computer Science Fudan University China Shanghai Key Laboratory of Intelligent Information Processing Fudan University China College of Computer Science and Technology Zhejiang University China

ISBN: (纸本)9798891761681

The training process of large language models (LLMs) often involves varying degrees of test data contamination (Yang et al., 2023b).Although current LLMs are achieving increasingly better performance on various benchmarks, their performance in practical applications does not always match their benchmark *** of benchmarks can prevent the accurate assessment of LLMs' true ***, constructing new benchmarks is costly, labor-intensive and still carries the risk of ***, in this paper, we ask the question "Can we reuse these leaked benchmarks for LLM evaluation?" We propose Inference-Time Decontamination (ITD) to address this issue by detecting and rewriting leaked samples without altering their *** can mitigate performance inflation caused by memorizing leaked *** proof-of-concept experiments demonstrate that ITD reduces inflated accuracy by 22.9% on GSM8K and 19.0% on *** MMLU, using Inference-time Decontamination can lead to a decrease in the results of Phi3 and Mistral by 6.7% and 3.6% *** hope that ITD can provide more truthful evaluation results for large language models. © 2024 Association for Computational Linguistics.

关键词： Decontamination

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：