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

CACP: Covariance-Aware Cross-Domain Prototypes for Domain Adaptive Semantic Segmentation

引用

IEEE Transactions on Multimedia 2025年

作者： Xue, Yanbing Tian, Xinyu Zhang, Feifei Wen, Xianbin Gao, Zan Chen, Shengyong Tianjin University of Technology Key Laboratory of Computer Vision and System Ministry of Education Tianjin300382 China

Domain adaptive semantic segmentation aims to reduce domain shifts / discrepancies between source and target domains, improving the source domain model's generalization ability to the target domain. Recently, prototypical methods, which primarily use single-source or single-target domain prototypes as category centers to aggregate features from both domains, have achieved competitive performance in this task. However, due to large domain shifts, single-source domain prototypes have finite generalization ability and not all source domain knowledge is conducive to model generalization. Single-target domain prototypes are noisy because they are prematurely initialized with all features filtered by pseudo labels, which causes error accumulation in the prototypes. To address these issues, we proposes a covariance-aware cross-domain prototypes method (CACP) to achieve robust domain adaptation. We propose to use both domain prototypes to dynamically rectify pseudo labels in the target domain, effectively reducing the recognition difficulty of hard target domain samples and narrowing the gap between features of the same category in both domains. In addition, to further generalize the model to the target domain, we propose two modules based on covariance correlation, FSPC (Features Selection by Prototypes Covariances) and WSPC (Weighting Source by Prototypes Coefficients), to learn discriminative characteristics. FSPC selects highly correlated features to update target domain prototypes online, denoising and enhancing discriminativeness between categories. WSPC utilizes the correlation coefficients between target domain prototypes and source domain features to weight each point in the source domain, eliminating the information interference from the source domain. In particular, CACP achieves excellent performance on the GTA5 \to Cityscapes and SYNTHIA \to Cityscapes tasks with minimal computational resources and time. © 1999-2012 IEEE.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Foreground Prediction for Image Composition with Local and Global Feature Fusion 16

Foreground Prediction for Image Composition with Local and G...

引用

2024 16th International Conference on Graphics and Image Processing, ICGIP 2024

作者： Sun, Liliang He, Yuanlie Li, Wensheng Feng, Fujian Liang, Yihui School of Computer Guangdong University of Technology Guangzhou510000 China School of Computer Science Zhongshan Institute University of Electronic Science and Technology of China Zhongshan528400 China Guizhou Key Laboratory of Pattern Recognition and Intelligent System Guizhou Minzu University Guiyang550025 China

ISBN: (数字)9781510688780

ISBN: (纸本)9781510688773

This paper focuses on the image composition of transparent objects, where existing image matting methods suffer from composition errors due to the lack of accurate foreground during the composition process. We propose a foreground prediction model named ALGM, which leverages the local feature extraction capabilities of Convolutional Neural Networks (CNNs) and incorporates an attention mechanism for global information modeling. The proposed alpha-assisted foreground prediction module extracts foreground information from the original image and conveys it. The extracted foreground color information is combined with the deep structural features of the encoder and used for foreground color prediction. ALGM reduces image composition errors in the quantitative data from the Composition-1k dataset and improves the visual quality of composed images on the AIM-500 and Transparent-460 datasets. © 2025 SPIE.

关键词： Prediction models

来源：评论

学校读者我要写书评

暂无评论

Infrared Small Target Detection with Feature Refinement and Context Enhancement 31st

Infrared Small Target Detection with Feature Refinement and...

引用

31st International Conference on Multimedia Modeling, MMM 2025

作者： Li, Xiuhong Zhu, Xinyue Li, Boyuan Li, Songlin Wang, Luyao Jia, Zhenhong School of Computer Science and Technology Xinjiang University Urumgi China Key Laboratory of Signal Detection and Processing Xinjiang University Urumqi China College of Mathematics and System Science Xinjiang University Urumqi China

ISBN: (纸本)9789819620609

Infrared small target detection has received widespread application and attention in both civilian and military fields. However, due to the very small size and lack of unique features of these targets, existing methods often suffer from inaccurate edge localization, and the targets are easily overwhelmed by complex backgrounds. To effectively address these issues, we designed a Feature Refinement and Context Enhancement Network (FCNet). The network consists of a multi-branch feature extraction module (MFEM), which integrates ordinary convolution and center difference convolution for diversified feature extraction, and further refines features through attention mechanism to enhance feature expression ability. In addition, in order to better capture the contextual information of the target, we have introduced the Context Enhancement Module (CEM). The CEM improves the robustness and detection accuracy of the network in complex backgrounds by preserving and enhancing important information in the image. Especially when processing infrared images with complex backgrounds and low signal-to-noise ratios, CEM can effectively highlight the target area and reduce false positives and false negatives. The experimental results on the SIRST dataset show that our FCNet performs well on multiple evaluation metrics. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Infrared imaging

来源：评论

学校读者我要写书评

暂无评论

Non-Line-of-Sight Multipath Classification Method for BDS Using Convolutional Sparse Autoencoder with LSTM

引用

Tsinghua Science and technology 2025年第1期30卷 68-86页

作者： Yahang Qin Zhenni Li Shengli Xie Bo Li Ming Liu Victor Kuzin School of Automation Guangdong University of Technologyand also with Guangdong-Hong Kong-Macao Joint Laboratory for Smart Discrete ManufacturingGuangzhou 510006China School of Automation Guangdong University of TechnologyGuangdong Key Laboratory of IoT Information TechnologyGuangzhou 510006China 111 Center for Intelligent Batch Manufacturing Based on IoT Technology Key Laboratory of Intelligent Information Processing and System Integration of IoTMinistry of EducationGuangzhou 510006China Department of Electronic and Computer Engineering Hong Kong University of Science and TechnologHong Kong 999077China Academician of Russian Engineering Academy Moscow Moscow 125009Russia

Multipath signal recognition is crucial to the ability to provide high-precision absolute-position services by the BeiDou Navigation Satellite system(BDS).However,most existing approaches to this issue involve supervised machine learning(ML)methods,and it is difficult to move to unsupervised multipath signal recognition because of the limitations in signal *** by an autoencoder with powerful unsupervised feature extraction,we propose a new deep learning(DL)model for BDS signal recognition that places a long short-term memory(LSTM)module in series with a convolutional sparse autoencoder to create a new autoencoder ***,we propose to capture the temporal correlations in long-duration BeiDou satellite time-series signals by using the LSTM module to mine the temporal change patterns in the time ***,we develop a convolutional sparse autoencoder method that learns a compressed representation of the input data,which then enables downscaled and unsupervised feature extraction from long-duration BeiDou satellite series ***,we add an l_(1/2) regularizer to the objective function of our DL model to remove redundant neurons from the neural network while ensuring recognition *** tested our proposed approach on a real urban canyon dataset,and the results demonstrated that our algorithm could achieve better classification performance than two ML-based methods(e.g.,11%better than a support vector machine)and two existing DL-based methods(e.g.,7.26%better than convolutional neural networks).

关键词： convolutional sparse autoencoder BeiDou Navigation Satellite system(BDS) long short-term memory(LSTM) multipath classification

来源：评论

学校读者我要写书评

暂无评论

React Agent-Based Question Answering Method for Biology in National College Entrance Examination 24

React Agent-Based Question Answering Method for Biology in N...

引用

5th International Symposium on Artificial Intelligence for Medicine Science, ISAIMS 2024

作者： Chen, Yulin Luo, Jing Tu, Xinhui College of Computer Science and Technology Wuhan University of Science and Technology Hubei Wuhan China Hubei Province Key Laboratory of Intelligent Information Processing and Real-time Industrial System Hubei Wuhan China College of Computer Science Central China Normal University Hubei Wuhan China

ISBN: (纸本)9798400717826

Recent advancements in large language models have led to significant improvements in various natural language processing tasks, including automated question answering. However, these models still struggle with providing accurate responses to complex biology questions, such as those found in the National College Entrance Examination. To address this issue, researchers have explored retrieval-augmented generation techniques, which incorporate external knowledge sources to enhance answer accuracy. While these methods have shown promise in improving efficiency, they often lack the ability to perform in-depth analysis of complex questions. In response to these limitations, the paper proposes a novel ReAct agent-based question answering method. This method combines biology textbook document libraries with Google search engine capabilities, leveraging the strengths of large language models. By integrating these components, this method can engage in autonomous decision-making and multi-step reasoning, allowing for a more thorough analysis of complex biology questions. Experimental results demonstrate the effectiveness of this new approach. When compared to responses generated directly by ChatGPT and GPT-4o models, the method showed significant improvements in accuracy for biology questions from the National College Entrance Examination. Specifically, the new method achieved accuracy increases of 11.1% and 4.64% over ChatGPT and GPT-4o, respectively. © 2024 Copyright held by the owner/author(s).

关键词： Question answering

来源：评论

学校读者我要写书评

暂无评论

DSAA: Cross-modal transferable double sparse adversarial attacks from images to videos

引用

Neurocomputing 2025年 639卷

作者： Han, Xueqiang Zhang, Shihui Wang, Houlin Tian, Qing School of Information Science and Engineering Yanshan University Hebei Qinhuangdao China The Key Laboratory for Computer Virtual Technology and System Integration of Hebei Province Hebei Qinhuangdao China Department of Computer Science University of Alabama at Birmingham AL United States

Recent preliminary research has examined cross-modal attacks, enabling the transition of unlabeled attacks from images to videos. The current cross-modal attacks are mainly dense attacks, while sparse attacks have not been explored. To bridge this gap, this paper proposes a cross-modal transferable double sparse adversarial attack (DSAA) method. Specifically, DSAA incorporates both spatial sparsity and temporal sparsity. Firstly, we design a novel double-cross-shaped patch paradigm to achieve spatial sparsity. Leveraging this patch for perturbation addition enables a more effective global impact on video frames. Subsequently, we propose a uniform temporal sparsity strategy to achieve temporal sparsity. Even if only half of the frames are perturbed, the resulting average attack success rate remains comparable to that of perturbing all frames. Furthermore, we take into account the luminance, contrast, and structure of images and propose a structural similarity loss function (SSLF). By utilizing SSLF to measure the similarity of intermediate feature maps and incorporating it into perturbation optimization, DSAA achieves excellent attack performance. Extensive experiments demonstrate DSAA's superiority over state-of-the-art approaches in terms of attack success rate. The proposed DSAA uses video adversarial examples generated by four ImageNet-pretrained image models on Kinetics-400 and UCF-101 to attack six video recognition models, achieving average attack success rates of 85.83% and 82.02%, respectively. © 2025 Elsevier B.V.

关键词： Cross-modal attacks Spatial sparsity Temporal sparsity Unlabeled attacks Video adversarial examples

来源：评论

学校读者我要写书评

暂无评论

Refining Boundary Sample Space for Differentiable Rendering 16

Refining Boundary Sample Space for Differentiable Rendering

引用

2024 16th International Conference on Graphics and Image Processing, ICGIP 2024

作者： Gao, Ruicheng Qi, Yue State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and Engineering Beihang University Beijing100191 China Jingdezhen Research Institute Beihang University Jingdezhen333000 China Qingdao Research Institute Beihang University Qingdao266104 China

ISBN: (数字)9781510688780

ISBN: (纸本)9781510688773

The objective of physics-based differentiable rendering (PBDR) is to propagate gradients between scene parameters and the intensities of image pixels in a manner that is physically correct. The gradients obtained can be applied in optimization algorithms for the reconstruction of 3D geometry or materials, or they can be further propagated into neural network to learn neural representations of the scene. However, applying automatic differentiation techniques directly to the primary rendering process will result in biased gradients, as the rendering integral contains moving high-dimensional discontinuities. Based on how these discontinuities are managed-either implicitly or explicitly-existing PBDR methods can be categorized into two groups: reparameterization methods and boundary sampling methods. Boundary sampling methods need to construct paths that have one segment tangent to the geometry being differentiated in order to estimate a boundary integral to address the discontinuities explicitly. Such paths are usually constructed by sampling the tangent segment first and then extending it to complete the paths for subsequent processing. Fortunately, the number of dimensions in the space composed of such tangent segments is only three. In scenes comprised solely of triangle meshes, the first dimension is used to parameterize all the edges on the mesh, which determines a point on the tangent segment. The remaining two dimensions are used to parameterize the direction of the tangent segment. However, state-of-the-art boundary sampling methods parameterize the first dimension uniformly, which is inefficient because only a small portion of the edges contributes to the boundary integral, resulting in wasted parameter space. In this paper, we parameterize the first dimension by considering both edge length and contributions, thereby allocating more parameter space to important edges. Experiments demonstrate that our methods achieve lower variance gradients in the forward dif

关键词： Mesh generation

来源：评论

学校读者我要写书评

暂无评论

Binary-Encoding-Based Quantized Kalman Filter: An Approximate MMSE Approach

引用

IEEE Transactions on Automatic Control 2025年第5期70卷 3181-3196页

作者： Liu, Qinyuan Nie, Yao Wang, Zidong Dong, Hongli Jiang, Changjun Tongji University School of Computer Science and Technology Shanghai201804 China Tongji University Key Laboratory of Embedded System and Service Computing Ministry of Education Shanghai201804 China Tongji University Shanghai Artificial Intelligence Laboratory Shanghai201804 China Brunel University London Department of Computer Science UxbridgeUB8 3PH United Kingdom Northeast Petroleum University Artificial Intelligence Energy Research Institute Daqing163318 China Northeast Petroleum University Heilongjiang Provincial Key Laboratory of Networking and Intelligent Control Daqing163318 China

In this article, the Kalman filter design problem is investigated for linear discrete-time systems under binary encoding schemes. Under such a scheme, the local information is quantized into a bit string by the remote sensor based on a probabilistic quantizer, and then the bit string is transmitted via memoryless binary symmetric channels (BSCs). Due to the communication link noises, the bit flipping occurs in a random manner, and thus, the transmission of the bit string would suffer from specific bit-error rates. With the received bits, a recursive binary-encoding-based quantized Kalman filter is established in the approximate minimum mean-square error (MMSE) sense, which relies on the Gaussian approximation of the conditional probability density function at each iteration. Furthermore, the proposed estimator is shown to be of a Kalman-like type through performance analysis, which exhibits computational complexity comparable to the conventional Kalman filter. Subsequently, a posterior Cramér-Rao lower bound is derived for the proposed binary-encoding-based quantized Kalman filter. The effectiveness of the proposed estimator is demonstrated through numerical results. © 1963-2012 IEEE.

关键词： Mean square error

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

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：