检索结果-内蒙古大学图书馆

International Joint Conference on Neural networks (IJCNN)

作者： Xinrui Zhang Xudong Luo Jiaye Wu Guangxi Key Lab of Multi-Source Information Mining & Security School of Computer Science and Engineering Guangxi Normal University Guilin China

In legal practice, judicial professionals often need to extract useful information from numerous legal documents. The key to legal information extraction is the Named Entity Recognition (NER) of legal documents. To address three critical issues in NER of legal documents, this paper proposes a RoBERTa-GlobalPointer-based method for NER of legal documents. Specifically, we first use RoBERTa (a variant of the pre-trained language model BERT) to extract char-level feature representations of a legal document, and use the Skip-Gram method to extract its word-level feature representations, and fuse them to better capture the contextual information of entities in the document. Then, according to the concatenated result, we use the GlobalPointer method to calculate the score of each subsequence of the document, to which it is an entity of a certain type. Finally, we employ the balanced softmax function to determine whether or not a subsequence of the document is an entity of a certain type according to its score calculated by GlobalPointer. Our evaluation experiments on the Chinese judicial domain dataset show that the proposed method outperforms the state-of-the-art baseline methods.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Adfq-Vit: Activation-Distribution-Friendly Post-Training Quantization for Vision Transformers

SSRN

引用

SSRN 2024年

作者： Jiang, Yanfeng Sun, Ning Xie, Xueshuo Yang, Fei Li, Tao College of Computer Science Nankai University Tianjin China Tianjin Key Laboratory of Network and Data Security Technology Tianjin China Zhejiang Lab Zhejiang Hangzhou China Haihe Lab of ITAI Tianjin China

Vision Transformers (ViTs) have exhibited exceptional performance across diverse computer vision tasks, while their substantial parameter size incurs significantly increased memory and computational demands, impeding effective inference on resource-constrained devices. Quantization has emerged as a promising solution to mitigate these challenges, yet existing methods still suffer from significant accuracy loss at low-bit. We attribute this issue to the distinctive distributions of post-LayerNorm and post-GELU activations within ViTs, rendering conventional hardware-friendly quantizers ineffective, particularly in low-bit scenarios. To address this issue, we propose a novel framework called Activation-Distribution-Friendly post-training Quantization for Vision Transformers, ADFQ-ViT. Concretely, we introduce the Per-Patch Outlier-aware Quantizer to tackle irregular outliers in post-LayerNorm activations. This quantizer refines the granularity of the uniform quantizer to a per-patch level while retaining a minimal subset of values exceeding a threshold at full-precision. To handle the non-uniform distributions of post-GELU activations between positive and negative regions, we design the Shift-Log2 Quantizer, which shifts all elements to the positive region and then applies log2 quantization. Moreover, we present the Attention-score enhanced Module-wise Optimization which adjusts the parameters of each quantizer by reconstructing errors to further mitigate quantization error. Extensive experiments demonstrate ADFQ-ViT provides significant improvements over various baselines in image classification, object detection, and instance segmentation tasks at 4-bit. Specifically, when quantizing the ViT-B model to 4-bit, we achieve a 10.23% improvement in Top-1 accuracy on the ImageNet dataset. © 2024, The Authors. All rights reserved.

关键词： Chemical activation

来源：评论

学校读者我要写书评

暂无评论

Unsupervised Hyperspectral Pansharpening via Low-rank Diffusion Model

arXiv

引用

arXiv 2023年

作者： Rui, Xiangyu Cao, Xiangyong Pang, Li Zhu, Zeyu Yue, Zongsheng Meng, Deyu Xi’an Jiaotong University Xianning West Road Shaanxi Xi’an710049 China The School of Mathematics and Statistics Ministry of Education Key Lab of Intelligent Network Security China The School of Computer Science and Technology Ministry of Education Key Lab For Intelligent Networks and Network Security China The College of Artificial Intelligence

Hyperspectral pansharpening is a process of merging a high-resolution panchromatic (PAN) image and a low-resolution hyperspectral (LRHS) image to create a single high-resolution hyperspectral (HRHS) image. Existing Bayesian-based HS pansharpening methods require designing handcraft image prior to characterize the image features, and deep learning-based HS pansharpening methods usually require a large number of paired training data and suffer from poor generalization ability. To address these issues, in this work, we propose a low-rank diffusion model for hyperspectral pansharpening by simultaneously leveraging the power of the pre-trained deep diffusion model and better generalization ability of Bayesian methods. Specifically, we assume that the HRHS image can be recovered from the product of two low-rank tensors, i.e., the base tensor and the coefficient matrix. The base tensor lies on the image field and has a low spectral dimension. Thus, we can conveniently utilize a pre-trained remote sensing diffusion model to capture its image structures. Additionally, we derive a simple yet quite effective way to pre-estimate the coefficient matrix from the observed LRHS image, which preserves the spectral information of the HRHS. Experimental results demonstrate that the proposed method performs better than some popular traditional approaches and gains better generalization ability than some DL-based methods. The code is released in https://***/xyrui/PLRDiff. Copyright © 2023, The Authors. All rights reserved.

关键词： Tensors

来源：评论

学校读者我要写书评

暂无评论

A Two-Step Method for Dynamics of Abstract Argumentation 1

引用

3rd International Conference on Distributed Artificial Intelligence, DAI 2021

作者： Jing, Xiaoxin Luo, Xudong College of Political Science and Law Capital Normal University Beijing China Guangxi Key Lab of Multi-Source Information Mining and Security School of Computer Science and Engineering Guangxi Normal University Guilin China

ISBN: (数字)9783030946623

ISBN: (纸本)9783030946616

This paper proposes a two-step method to handle the dynamics of argumentation. Firstly, we reach the conflict-free labellings of the updated argumentation framework by the intersection of expansions of the conflict-free labellings of argumentation frameworks before updating. Then we select the conflict-free labellings which have the least illegally labelled arguments when restricted to part of both argumentation frameworks in the update process. Finally, we prove the soundness and completeness of our method under the complete semantics. In other words, the complete labellings of the resulted argumentation framework after update is the same as that of our method. © 2022, Springer Nature Switzerland AG.

关键词：

来源：评论

学校读者我要写书评

暂无评论

CA-UAP: Content-Agnostic Universal Adversarial Perturbation for Enhanced Generalization

CA-UAP: Content-Agnostic Universal Adversarial Perturbation ...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Rui Lu Ziqiang He Jingyang Wen Xiangui Kang Z. Jane Wang Guangdong Key Lab of Information Security School of Computer Science and Engineering Sun Yat-Sen University Guangzhou China College of Electrical and Information Engineering North Minzu University Yinchuan China Department of Electrical and Computer Engineering University of British Columbia Canada

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

Deep Neural networks (DNNs) have been shown vulnerable to universal adversarial perturbation (UAP), which are imperceptible and capable of fooling the target model for most samples. Existing universal attack methods mainly focus on aggregating the gradient obtained from global image features to directly optimize (noise-based) or indirectly generate (generator-based) UAP. However, such methods do not yet consider improving the generalization of UAP from the perspective of making the perturbation irrelevant to image content. We note that minimizing self-similarity is helpful to make the UAP irrelevant to the image content. Therefore, we propose a novel Content-Agnostic UAP (CA-UAP), which combines global image features and local patch features to optimize UAP. Specifically, we introduce a self-similarity loss that encourages minimizing the similarity between adversarial perturbed global images and their randomly cropped local regions, making the UAP agnostic to image content and consequently enhancing UAP generalization. Extensive experiments on the ILSVRC 2012 dataset demonstrate that our proposed method outperforms existing methods in both untargeted and targeted attacks, e.g., improving the average fooling rate from 79.12% (achieved by the state-of-the-art method) to 82.25% in targeted attacks.

关键词： Correlation Perturbation methods Buildings Artificial neural networks Signal processing Transformers Filling Acoustics Speech processing

来源：评论

学校读者我要写书评

暂无评论

PGD-Imp: Rethinking and Unleashing Potential of Classic PGD with Dual Strategies for Imperceptible Adversarial Attacks

arXiv

引用

arXiv 2024年

作者： Li, Jin Yu, Zitong He, Ziqiang Wang, Z. Jane Kang, Xiangui Guangdong Key Lab of Information Security School of Computer Science and Engineering Sun Yat-Sen University China School of Computing and Information Technology Great Bay University China Electrical and Computer Engineering Dept University of British Columbia Canada

Imperceptible adversarial attacks have recently attracted increasing research interests. Existing methods typically incorporate external modules or loss terms other than a simple lp-norm into the attack process to achieve imperceptibility, while we argue that such additional designs may not be necessary. In this paper, we rethink the essence of imperceptible attacks and propose two simple yet effective strategies to unleash the potential of PGD, the common and classical attack, for imperceptibility from an optimization perspective. Specifically, the Dynamic Step Size is introduced to find the optimal solution with minimal attack cost towards the decision boundary of the attacked model, and the Adaptive Early Stop strategy is adopted to reduce the redundant strength of adversarial perturbations to the minimum level. The proposed PGD-Imperceptible (PGD-Imp) attack achieves state-of-the-art results in imperceptible adversarial attacks for both untargeted and targeted scenarios. When performing untargeted attacks against ResNet-50, PGD-Imp attains 100% (+0.3%) ASR, 0.89 (-1.76) l2 distance, and 52.93 (+9.2) PSNR with 57s (-371s) running time, significantly outperforming existing methods. Copyright © 2024, The Authors. All rights reserved.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

High-quality Arabidopsis thaliana Genome Assembly with Nanopore and HiFi Long Reads

引用

Genomics, Proteomics & Bioinformatics 2022年第1期20卷 4-13页

作者： Bo Wang Xiaofei Yang Yanyan Jia Yu Xu Peng Jia Ningxin Dang Songbo Wang Tun Xu Xixi Zhao Shenghan Gao Quanbin Dong Kai Ye MOE Key Laboratory for Intelligent Networks&Network Security Faculty of Electronic and Information EngineeringXi’an Jiaotong UniversityXi’an 710049China School of Computer Science and Technology Faculty of Electronic and Information EngineeringXi’an Jiaotong UniversityXi’an 710049China School of Life Science and Technology Xi’an Jiaotong UniversityXi’an 710049China School of Automation Science and Engineering Faculty of Electronic and Information EngineeringXi’an Jiaotong UniversityXi’an 710049China Genome Institute the First Affiliated Hospital of Xi’an Jiaotong UniversityXi’an 710061China

Arabidopsis thaliana is an important and long-established model species for plant molecular biology,genetics,epigenetics,and ***,the latest version of reference genome still contains a significant number of missing ***,we reported a high-quality and almost complete Col-0 genome assembly with two gaps(named Col-XJTU)by combining the Oxford Nanopore Technologies ultra-long reads,Pacific Biosciences high-fidelity long reads,and Hi-C *** total genome assembly size is 133,725,193 bp,introducing 14.6 Mb of novel sequences compared to the TAIR10.1 reference *** five chromosomes of the Col-XJTU assembly are highly accurate with consensus quality(QV)scores>60(ranging from 62 to 68),which are higher than those of the TAIR10.1 reference(ranging from 45 to 52).We completely resolved chromosome(Chr)3 and Chr5 in a telomere-to-telomere ***4 was completely resolved except the nucleolar organizing regions,which comprise long repetitive DNA *** Chrl centromere(CEN1),reportedly around 9 Mb in length,is particularly challenging to assemble due to the presence of tens of thousands of CEN180 satellite *** the cutting-edge sequencing data and novel computational approaches,we assembled a 3.8-Mb-long CEN1 and a 3.5-Mb-long *** also investigated the structure and epigenetics of *** clusters of CEN180 monomers were detected,and the centromere-specific histone H3-like protein(CENH3)exhibited a strong preference for CEN180 Cluster ***,we observed hypomethylation patterns in CENH3-enriched *** believe that this high-quality genome assembly,Col-XJTU,would serve as a valuable reference to better understand the global pattern of centromeric polymorphisms,as well as the genetic and epigenetic features in plants.

关键词： Centromere architecture CENH3 Bacterial artificial chromosome Telomere-to-telomere Model plant

来源：评论

学校读者我要写书评

暂无评论

Self-Attention Policy Optimization for Task Offloading and Resource Allocation in Low-Carbon Agricultural Consumer Electronic Devices

引用

IEEE Transactions on Consumer Electronics 2025年

作者： Huang, Yi Zeng, Jisong Wei, Yanting Chen, Miaojiang Xiao, Wenjing Yang, Yang Liu, Zhiquan Farouk, Ahmed Song, Houbing Herbert Guangxi University School of Computer Electronics and Information China Guangxi Key Laboratory of Multimedia Communications and Network Technology Nanning530004 China Jinan University College of Cyber Security Guangzhou510632 China Faculty of Computers and Artificial Intelligence Egypt Department of Information Systems BaltimoreMD21250 United States

In recent years, the widespread use of edge agricultural consumer electronics has greatly contributed to the level of intelligence in agricultural production, bringing higher efficiency and quality. However, offloading all tasks to the cloud incurs significant latency and resource waste, while relying solely on edge computing fails to meet the computational demands of the entire system. To solve the above problems, we introduce the device-edge-cloud (DEC) three-layer architecture, where agri-consumer electronics devices can partially offload tasks to the edge, and the edge can partially offload tasks to the cloud, i.e., agri-consumer electronics can realize device-edge-cloud collaborative computation. Second, we model the joint computation offloading and resource allocation optimization problem as a non-convex optimization and propose a novel Self-Attention Policy Optimization (SAPO) algorithm to solve it. Experiments show that the joint optimization performance of the proposed SAPO exceeds the baseline, and it is suitable for many different models. Compared with fully connected networks, it has better convergence and robustness, with a convergence speed 50% faster than the fully connected networks. The proposed SAPO algorithm has good scalability and adaptability, and has the potential to be extended to smart agricultural computing scenarios with non-convex optimization. © 1975-2011 IEEE.

关键词： Computation offloading

来源：评论

学校读者我要写书评

暂无评论

A novel violent video detection method based on improved C3D and transfer learning 3

A novel violent video detection method based on improved C3D...

引用

2022 3rd International Conference on computer information and Big Data Applications, CIBDA 2022

作者： Liu, Guoqing Wang, Zhiwen Zhang, Haipeng Guo, Xin Wang, Yuhang Zhang, Canlong College of electrical electronics and computer Science Guangxi University of Science and Technology LiuZhou545006 China Guangxi Key Lab of Multi-source Information Mining & Security Guangxi Normal University Guilin541004 China

ISBN: (纸本)9783800758760

In recent years, due to the widespread use of surveillance systems, the crime rates significantly decreased and public security is guaranteed. However, instead of stopping ongoing crimes, these surveillance videos is often to provide clues and evidence after crimes occur. Manual monitoring is time-consuming and laborious, so it is necessary to automatically detect these violent behaviors in videos. In this paper, we propose a new violent video detection method, which improves C3D so that it can detect violent videos more effectively. In addition, we sample the input video during training, extracting one frame every two frames. In the test, the input video is not sampled. We also pre-train part of the network on the Sports-1M dataset using transfer learning method. The experimental results show that the accuracy of the method proposed in this paper reaches 99.5% on Hockey Fights, 98.37% on Violent Crowds, and 100% on Action Movie, which is very simple and effective. © VDE VERLAG GMBH - Berlin - Offenbach.

关键词： Crime

来源：评论

学校读者我要写书评

暂无评论

Towards an Ontology of Traceable Impact Management in the Food Supply Chain

arXiv

引用

arXiv 2025年

作者： Gajderowicz, Bart Fox, Mark S. Gao, Yongchao Urban Data Centre School of Cities University of Toronto Toronto Canada Key Laboratory of Computing Power Network and Information Security Shandong Computer Science Center Qilu University of Technology Shandong Academy of Sciences Jinan China

The pursuit of quality improvements and accountability in the food supply chains, especially how they relate to food-related outcomes, such as hunger, has become increasingly vital, necessitating a comprehensive approach that encompasses product quality and its impact on various stakeholders and their communities. Such an approach offers numerous benefits in increasing product quality and eliminating superfluous measurements while appraising and alleviating the broader societal and environmental repercussions. A traceable impact management model (TIMM) provides an impact structure and a reporting mechanism that identifies each stakeholder’s role in the total impact of food production and consumption stages. The model aims to increase traceability’s utility in understanding the impact of changes on communities affected by food production and consumption, aligning with current and future government requirements, and addressing the needs of communities and consumers. This holistic approach is further supported by an ontological model that forms the logical foundation and a unified terminology. By proposing a holistic and integrated solution across multiple stakeholders, the model emphasizes quality and the extensive impact of championing accountability, sustainability, and responsible practices with global traceability. With these combined efforts, the food supply chain moves toward a global tracking and tracing process that not only ensures product quality but also addresses its impact on a broader scale, fostering accountability, sustainability, and responsible food production and consumption. © 2025, CC BY.

关键词：

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：