检索结果-内蒙古大学图书馆

31st International Conference on Computational Linguistics, COLING 2025

作者： Wei, Kaiwen Zhong, Jiang Zhang, Hongzhi Zhang, Di Jin, Li Yu, Yue Zhang, Jingyuan College of Computer Science Chongqing University China Kuaishou Inc. China Key Laboratory of Network Information System Technology Aerospace Information Research Institute Chinese Academy of Sciences China

ISBN: (纸本)9798891761964

Large Language Models (LLMs) exhibit remarkable generative capabilities, enabling the generation of valuable information. Despite these advancements, previous research found that LLMs sometimes struggle with adhering to specific constraints (e.g., in specific place or at specific time), at times even overlooking them, which leads to responses that are either too generic or not fully satisfactory. Existing approaches attempted to address this issue by decomposing, rewriting input instructions, or reflecting on prior failings, yet they fall short in adequately emphasizing specific constraints and in unlocking the underlying knowledge (e.g., programming within the context of software development). In response, this paper proposes a simple yet effective method Chain-of-Specificity (CoS). Specifically, CoS iteratively emphasizes the specific constraints in the input instructions, unlocks knowledge within LLMs, and refines responses. Experiments conducted on publicly available and self-build complex datasets demonstrate that CoS outperforms existing methods in enhancing generated content especially for the specificity. Besides, as the number of specific constraints increase, other baselines falter, while CoS still performs well. Moreover, we show that distilling responses generated by CoS effectively enhances the ability of smaller models to follow the constrained instructions. © 2025 Association for Computational Linguistics.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

DSDIR: A Two-Stage Method for Addressing Noisy Long-Tailed Problems in Malicious Traffic Detection

DSDIR: A Two-Stage Method for Addressing Noisy Long-Tailed P...

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Li, Guoliang Zhang, Ruiqi Sun, Zhe Xing, Lingkai Zhang, Yu College of Cyber Science Nankai University Tianjin China College of Computer Science Nankai University Tianjin China Tianjin Key Laboratory of Network and Data Security Technology Tianjin China Key Laboratory of Data and Intelligent System Security Ministry of Education Tianjin China

ISBN: (纸本)9798350368741

In recent years, deep learning based malicious traffic detection (MTD) systems have demonstrated remarkable success. However, their effectiveness tend to decrease because most malicious traffic datasets are suffered from noisy-labeled and long-tailed problems. While numerous approaches have been developed to address these two problems individually, they become inefficient when confronting the combined challenge of noisy long-tailed data, as they typically tackle only a single adverse factor at a time. This paper proposes a two-stage method called Distribution-aware sample Selection and Dynamic Instance-based Relabeling (DSDIR), which simultaneously addresses the impacts of noisy-labeled and long-tailed problems. In the first stage, a noise-independent clean sample selection method is designed to obtain a clean dataset, which converts negative effects of the long-tailed problem into positive ones. In the second stage, dynamic instance-based relabeling is designed to train a model and improve the dataset's quality simultaneously. Eventually, DSDIR not only produces a balanced and noise-tolerant model but also obtains a clean dataset. Experimental results demonstrate that in the high noise condition of 60% and 80%, the accuracy rate of DSDIR is 5% higher than the state-of-the-art methods. Our code is available at https://***/nku-ligl/DSDIR. © 2025 IEEE.

关键词： deep learning distribution estimation dynamic relabeling malicious traffic detection noisy long-tailed problem

来源：评论

学校读者我要写书评

暂无评论

MCFTNet: Multimodal Cross-layer Fusion Transformer network for Hyperspectral and LiDAR Data Classification

引用

IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 2025年 18卷 12803-12818页

作者： Huang, Wei Wu, Tianren Zhang, Xueyu Li, Liangliang Lv, Ming Jia, Zhenhong Zhao, Xiaobin Ma, Hongbing Vivone, Gemine Xinjiang University School of Computer Science and Technology Urumqi830017 China Xinjiang University Key Laboratory of Signal Detection and Processing Urumqi830017 China Guangxi University School of Computer and Electronic Information Nanning530004 China Guangxi University Guangxi Key Laboratory of Multimedia Communications and Network Technology Nanning530004 China Beijing Institute of Technology School of Information and Electronics Beijing100081 China University of Science and Technology Beijing School of Computer and Communication Engineering Beijing100083 China Tsinghua University Department of Electronic Engineering Beijing100084 China National Research Council Tito85050 Italy NBFC National Biodiversity Future Center Palermo90133 Italy

Remote sensing image classification is a popular yet challenging field. Many researchers have combined convolutional neural networks (CNNs) and Transformers for hyperspectral image (HSI) classification tasks. However, in traditional Transformers, shallow-level information does not propagate well to deeper layers, which can lead to spatial variations and overfitting. Moreover, traditional Transformer models use an external classification (CLS) token that is randomly initialized and often struggles to generalize effectively. In this paper, we combine the strengths of HSI and light detection and ranging (LiDAR) data, using LiDAR as an external CLS token, which significantly enhances classification accuracy and reliability. We propose a new multimodal cross-layer fusion transformer network (MCFTNet), integrating CNNs with the latest Transformer networks. It includes a CNN for extracting spatial features and a hybrid cross-patch attention mechanism (HyCrossAttention) for land cover classification, leveraging LiDAR data to generate CLS and HSI patch tokens. More importantly, to reduce the loss of valuable information during the layer-by-layer propagation, we designed cross-layer skip connections. Through adaptive learning, these cross-layer fusions help address the gradient vanishing problem in deep networks while preserving early-layer features. This enables the model to better integrate information from different layers, enhancing both its stability and performance. We carried out in - depth experiments on commonly - used benchmark datasets, specifically, the University of Houston dataset, the Trento dataset, and the University of Southern Mississippi Gulf Park (MUUFL) dataset. We compared the results of the proposed MCFTNet model with those obtained from state-of-the-art Transformer models, classical CNNs, and traditional classifiers. As a result, the MCFTNet model outshone them all in terms of performance. © 2008-2012 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

A circuit area optimization of MK-3 S-box

引用

Cybersecurity 2025年第1期8卷 61-71页

作者： Yanjun Li Weiguo Zhang Yiping Lin Jian Zou Jian Liu Information Industry Information Security Evaluation Center The 15th Research Institute of China Electronic Technology Group CorporationBeijing 100083China Henan Key Laboratory of Network Cryptography Technology Zhengzhou 450001China Beijing Electronic Science and Technology Institute No.7 Fufeng RoadFengtai DistricBeijing 100083China College of Computer and Data Science Fuzhou UniversityFuzhou 350108China

In MILCOM 2015,Kelly et *** the authentication encryption algorithm MK-3,which applied the 16-bit *** paper aims to implement the 16-bit S-box with less circuit ***,we classifed the irreducible polynomials over F_(2n)into three *** we compared the logic gates required for multiplication over the fnite feld constructed by the three types of irreducible *** to the comparison result,we constructed the composite felds, F_((2^(4))^(2))and F_((2^(8))^(2)).Based on the isomorphism of fnite felds,the operations over F_(2)^(16)can be conducted over F_((2^(8))^(2)).Similarly,elements over F28 can be mapped to the corresponding elements over F_((2^(4)))^(2).Next,the SAT solver was used to optimize the operations over smaller feld F_(2)^(4).At last,the architecture of the optimized MK-3 S-box was worked *** with the implementation proposed by the original designer,the circuit area of the MK-3 S-box in this paper is reduced by at least 55.9%.

关键词： irreducible operations optimization

来源：评论

学校读者我要写书评

暂无评论

APRIL: Towards Scalable and Transferable Autonomous Penetration Testing in Large Action Space via Action Embedding

引用

IEEE Transactions on Dependable and Secure Computing 2025年第3期22卷 2443-2459页

作者： Zhou, Shicheng Liu, Jingju Lu, Yuliang Yang, Jiahai Hou, Dongdong Zhang, Yue Hu, Shulong National University of Defense Technology College of Electronic Engineering Hefei230037 China Anhui Province Key Laboratory of Cyberspace Security Situation Awareness and Evaluation Hefei230037 China Tsinghua University Institute for Network Sciences and Cyberspace Beijing National Research Center for Information Science and Technology Beijing100084 China National University of Defense Technology College of Computer Science and Technology Changsha410073 China

Penetration testing (pentesting) assesses cybersecurity through simulated attacks, while the conventional manual-based method is costly, time-consuming, and personnel-constrained. Reinforcement learning (RL) provides an agent-environment interaction learning paradigm, making it a promising way for autonomous pentesting. However, agents’ scalability in large action spaces and policy transferability across scenarios limit the applicability of RL-based autonomous pentesting. To address these challenges, we present a novel autonomous pentesting framework based on reinforcement learning (namely APRIL) to train agents that are scalable and transferable in large action spaces. In APRIL, we construct realistic, bounded, host-level state space via embedding techniques to avoid the complexities of dealing with unbounded network-level information. We employ semantic correlations between pentesting actions as prior knowledge to represent discrete action space into a continuous and semantically meaningful embedding space. Agents are then trained to reason over actions within the action embedding space, where two key methods are applied: an upper-confidence bound-based action refinement method to encourage efficient exploration, and a distance-aware loss to improve learning efficiency and generalization performance. We conduct experiments in simulated scenarios constructed based on virtualized vulnerable environments. The results demonstrate APRIL's scalability in large action spaces and its ability to facilitate policy transfer across diverse scenarios. © 2004-2012 IEEE.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

MDGCN-Lt:Fair Web API Classification with Sparse and Heterogeneous Data Based on Deep GCN

引用

清华大学学报自然科学版（英文版） 2025年第3期30卷 1294-1314页

作者： Boyuan Yan Yankun Zhang Wenwen Gong Haoyang Wan Wenwei Wang Weiyi Zhong Caixia Bu Department of School of Computer Science Qufu Normal UniversityRizhao 276500China Shandong Provincial University Laboratory for Protected Horticulture Weifang University of Science and TechnologyWeifang 261000China Department of Computer Science and Technology Tsinghua UniversityBeijing 100084China Shandong Provincial Key Laboratory of Network Based Intelligent Computing University of JinanJinan 250022China Department of Univalsoft Joint Stock Co. Ltd.Weifang 261000China

Developers integrate web Application Programming Interfaces(APIs)into edge applications,enabling data expansion to the edge computing area for comprehensive coverage of devices in that *** develop edge applications,developers search API categories to select APIs that meet specific ***,the accurate classification of APIs becomes critically ***,existing approaches,as evident on platforms like ***,face significant ***,sparsity in API data reduces classification accuracy in works focusing on single-dimensional API ***,the multidimensional and heterogeneous structure of web APIs adds complexity to data mining tasks,requiring sophisticated techniques for effective integration and analysis of diverse data ***,the long-tailed distribution of API data introduces biases,compromising the fairness of classification *** these challenges,we propose MDGCN-Lt,an API classification approach offering flexibility in using multi-dimensional heterogeneous *** tackles data sparsity through deep graph convolutional networks,exploring high-order feature interactions among API ***-Lt employs a loss function with logit adjustment,enhancing efficiency in handling long-tail data *** results affirm our approach's superiority over existing methods.

关键词： bidirectional encoder representations from transformers deep graph convolutional networks logit adjustment web application programming interface classification application programming interface correlation graph

来源：评论

学校读者我要写书评

暂无评论

Spatial-Temporal Semantic Feature Interaction network for Semantic Change Detection in Remote Sensing Images

引用

IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 2025年 18卷 12090-12102页

作者： Zhang, Yuhang Zhang, Wuxia Ding, Songtao Wu, Siyuan Lu, Xiaoqiang Xi'an University of Posts and Telecommunications Shaanxi Key Laboratory of Network Data Analysis and Intelligent Processing School of Computer Science and Technology Xi'an710121 China Xi'an University of Technology College of Computer Science and Engineering Shaanxi Xi'an710048 China Fuzhou University College of Physics and Information Engineering Fuzhou350002 China

Semantic Change Detection (SCD) in Remote Sensing Images (RSI) aims to identify changes in the type of Land Cover/Land Use (LCLU) corresponding to changed areas in RSI. The "from-to" information of the acquired image has more profound practical significance than Binary Change Detection (BCD). However, most deep learning-based SCD algorithms do not fully exploit the spatial-temporal information of multi-level features, leading to challenges in extracting LCLU features in complex changing scenes. To address these issues, we propose a Spatial-Temporal Semantic Feature Interaction network (STS-FINet) to improve the performance of SCD in RSI. The proposed STS-FINet comprises a Multi-Scale Feature Extraction Encoder (MS-FEE), a Transformer-based Multi-Level Feature Interaction module (TML-FI), and a Multi-Level Feature Fusion Decoder (ML-FFD). The MS-FEE extracts deep semantic and differential information from the RSI. The TML-FI is designed to mine the spatial-temporal information by extracting long-range dependencies and spatial information from multi-level features to improve spatial perception. Moreover, Mixed Spatial Reasoning Convolution block (MixSrc) is presented to enrich the spatial information by extracting the multi-scale features, thus improving the model's capability to interpret complex changing scenes. Finally, ML-FFD integrates the multi-level features, resulting in the generation of the semantic change map. The effectiveness of the proposed STS-FINet is verified on two high-resolution RSI datasets. Experimental results show that the proposed STS-FINet achieves better change detection performance than SOTA methods. © 2008-2012 IEEE.

关键词： Change detection

来源：评论

学校读者我要写书评

暂无评论

Complex networks with External Degree

引用

Chinese Journal of Electronics 2025年第3期23卷 442-447页

作者： Chunfang Li Lianzhong Liu Key Laboratory of Beijing Network Technology School of Computer Science and Engineering Beijing University of Aeronautics and Astronautics Beijing China

The evolvement of complex networks is based on self-contained ultra-large system, but in fact all possible described networks are almost not completeness sub-networks. We study the more complete information description of complex system through introducing autonomous system and external degree, and investigate the statistical properties of external edges. The external nodes, external edges, and external degree provide natural and powerful extension for the traditional network concept of complex physical system, capable of providing more comprehensive information about the connectivity properties of the graph topology. Complex network with external degree provides scalable scale analysis for complex system with as less cost as possible. Experiments reveal that external degree also follows power-law distribution.

关键词： Photography Visualization Costs Web services Education Complex networks network analyzers Topology Satellite images Complex systems

来源：评论

学校读者我要写书评

暂无评论

Disentangled Noisy Correspondence Learning

引用

IEEE Transactions on Image Processing 2025年 34卷 2602-2615页

作者： Dang, Zhuohang Luo, Minnan Wang, Jihong Jia, Chengyou Han, Haochen Wan, Herun Dai, Guang Chang, Xiaojun Wang, Jingdong Xi’an Jiaotong University School of Computer Science and Technology Ministry of Education Key Laboratory of Intelligent Networks and Network Security Shaanxi Province Key Laboratory of Big Data Knowledge Engineering Shaanxi Xi’an710049 China SGIT AI Laboratory Xi’an710048 China State Grid Corporation of China State Grid Shaanxi Electric Power Company Ltd. Xi’an710048 China University of Science and Technology of China School of Information Science and Technology Hefei230026 China Abu Dhabi United Arab Emirates Baidu Inc. Beijing100085 China

Cross-modal retrieval is crucial in understanding latent correspondences across modalities. However, existing methods implicitly assume well-matched training data, which is impractical as real-world data inevitably involves imperfect alignments, i.e., noisy correspondences. Although some works explore similarity-based strategies to address such noise, they suffer from sub-optimal similarity predictions influenced by modality-exclusive information (MEI), e.g., background noise in images and abstract definitions in texts. This issue arises as MEI is not shared across modalities, thus aligning it in training can markedly mislead similarity predictions. Moreover, although intuitive, directly applying previous cross-modal disentanglement methods suffers from limited noise tolerance and disentanglement efficacy. Inspired by the robustness of information bottlenecks against noise, we introduce DisNCL, a novel information-theoretic framework for feature Disentanglement in Noisy Correspondence Learning, to adaptively balance the extraction of modality-invariant information (MII) and MEI with certifiable optimal cross-modal disentanglement efficacy. DisNCL then enhances similarity predictions in modality-invariant subspace, thereby greatly boosting similarity-based alleviation strategy for noisy correspondences. Furthermore, DisNCL introduces soft matching targets to model noisy many-to-many relationships inherent in multi-modal inputs for noise-robust and accurate cross-modal alignment. Extensive experiments confirm DisNCL’s efficacy by 2% average recall improvement. Mutual information estimation and visualization results show that DisNCL learns meaningful MII/MEI subspaces, validating our theoretical analyses. © 1992-2012 IEEE.

关键词： Information theory

来源：评论

学校读者我要写书评

暂无评论

CurMIM: Curriculum Masked Image Modeling

CurMIM: Curriculum Masked Image Modeling

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Liu, Hao Wang, Kun Han, Yudong Wang, Haocong Hu, Yupeng Wang, Chunxiao Nie, Liqiang School of Software Shandong University Jinan China Key Laboratory of Computing Power Network and Information Security Ministry of Education Qilu University of Technology Shandong Academy of Sciences Jinan China School of Computer Science and Technology Harbin Institute of Technology Shenzhen China

ISBN: (纸本)9798350368741

Masked Image Modeling (MIM), following "mask-and-reconstruct" scheme, is a promising self-supervised method to learn scalable visual representation. Studies indicate that selecting an effective mask strategy is vital for MIM. However, existing approaches often rely on static pre-defined priors, which limit their ability to adapt mask strategies dynamically for network optimization. In this paper, we focus on the learning process of the network and introduce human-like curriculum into MIM for dynamic representation refinement, and propose an end-to-end framework Curriculum Masked Image Modeling (CurMIM). CurMIM consists of two components: Mask Priority Measurer, which acts as a curriculum learner to determine mask priority values using the network's intrinsic state information, and Dual Adaptive Selector, which serves as a curriculum scheduler to create effective masks based on these values. With negligible extra parameters, our curriculum-based method consistently establishes noticeable improvements across varying model sizes and benchmarks, showing effectiveness and generalization. © 2025 IEEE.

关键词： Curriculum Learning Image Representation Masked Image Modeling Visual Pre-training

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：