检索结果-内蒙古大学图书馆

21st Pacific Rim International Conference on Artificial Intelligence, PRICAI 2024

作者： Jiang, Zhihua Yang, Junzhan Rao, Dongning Department of Computer Science Jinan University Guangzhou510632 China School of Computer Guangdong University of Technology Guangzhou510006 China

ISBN: (纸本)9789819601189

Despite the impressive ability offered by pre-trained language models (PLMs) and large language models (LLMs), these models still face the challenges of the long-text summarization (LTS) task. On the one hand, due to the max input-length, PLMs adopt simple truncation or salience analysis for extracting part of sentences from the original input, leading to possible information loss. On the other hand, despite no limits on input length, the summaries generated by LLMs often introduce irrelevant information, leading to great gap between prediction and reference. To this end, this paper proposes a hybrid long-text summarization model, PLLM-TS (Leveraging PLMs and LLMs for Long-Text Summarization), which combines the advantages of both PLMs and LLMs. PLLM-TS is a pipeline composed of two modules. The first module is an extractive summarization model (e.g., BERTSum) enhanced with combined attention ideas. And the second module is a switching generative summarization model, dependent on the number of extracted sentences from the first module. Specifically, when the sentence number is bigger, we apply a PLM-based summarization model (e.g., Pegasus) to generate a summary from extracted sentences. By contrast, when the sentence number is smaller, we utilize a LLM-based summarization model (e.g., GPT-2) to rewrite extracted sentences into a shorter and more concise summary. We experiment with two public academic-paper datasets: arXiv and PubMed, showing the effectiveness and superiority of our method. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Modeling languages

来源：评论

学校读者我要写书评

暂无评论

PsSR: Hybrid Path Selection Mechanism for Efficient Image Super-Resolution

PsSR: Hybrid Path Selection Mechanism for Efficient Image Su...

引用

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

作者： Ding, Cheng Zhao, ZhongQiu Zhao, Yang Hefei University of Technology School of Computer Science and Information Engineering Hefei China

ISBN: (纸本)9798350368741

In practical applications, large images are processed in patches, many of which are simple and smooth, making them suitable for lighter network processing. This paper proposes a hybrid path selection mechanism that enables the dynamic routing of patches through the network, achieving a balance between accuracy and computational efficiency. The proposed method develops a dual-branch architecture, in which the two branches keep the same structural design, but differ in the number of channels. An efficient classification module is integrated for directing patches to the appropriate branches within the model. In addition, within each branch, we introduce an early exit strategy to further reduce the calculation cost. Subsequently, we enhance the restoration capabilities of the lightweight branch through channel and spatial feature distillation. Experimental results demonstrate that our Hybrid Path Selection based Super-Resolution (PsSR) mechanism effectively reduces computational costs while preserving the integrity of the backbone performance. © 2025 IEEE.

关键词： dual-branch knowledge distillation path selection super-resolution

来源：评论

学校读者我要写书评

暂无评论

Multi-scale Diffusion Features Fusion Network For Hyperspectral Image Classification 4

Multi-scale Diffusion Features Fusion Network For Hyperspect...

引用

4th Asia Conference on Information Engineering, ACIE 2024

作者： Tan, Xunyan Zhao, Zhuoyi Xu, Xiang Li, Si Guangdong University of Technology School of Computer Guangzhou China University of Electronic Science and Technology of China Zhongshan Institution Zhongshan China

ISBN: (纸本)9798350371925

Hyperspectral images (HSIs) contain rich spectral and spatial information, which provides valuable data foundation for the classification of ground objects. However, most of the existing classification methods are devoted to constructing discriminative models for feature extraction, and lack comprehensive interpretation of hyperspectral data, especially in the case of limited samples. Diffusion model, as a novel paradigm in generative model, has been widely concerned in various fields due to its powerful data modeling capabilities. In order to facilitate the application of diffusion features to hyperspectral image classification (HSIC), we propose a multi-scale diffusion features fusion network (dubbed 'MDFFNet'). The proposed MDFFNet adopts a hierarchical diffusion feature preprocessing and multi-scale feature fusion to efficiently learn the latent spectral-spatial feature representations from diffusion models, and then the fused diffusion features are used for subsequent classification. Experimental evaluations are carried out on two benchmark HSI datasets, and the results show impressive classification performance even with a limited number of training samples. © 2024 IEEE.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

LeakGAN-Based Causality Extraction in the Financial Field

LeakGAN-Based Causality Extraction in the Financial Field

引用

作者： Sun, Zhengyan Li, Xiaoqing Zhu, Guangli School of Computer Science and Engineering Anhui University of Science and Technology Huainan232001 China

Causality extraction model can quickly extract causality in text. It can be applied to event prediction, question-answering systems, and scenario generation. The traditional causality extraction pays more attention to the extraction of entities and ignores the deep semantic representation between cause entities and effect entities of the text. So the accuracy of causality extraction is low in the end. A causal relationship extraction model based on LeakGAN is proposed to solve this problem. The core task of the model is to analyze the causality existing in the review text (this paper focuses on explicit one-cause and one-effect analysis). It also realizes deep extraction under semantic enhancement. Four levels of causality extraction model based on LeakGAN are proposed: data preprocessing, pre-training, Bi-LSTM + Attention, and LeakGAN learning. Experimental results show that the model can improve the accuracy of causality extraction. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Extraction

来源：评论

学校读者我要写书评

暂无评论

MGTC: Multi-Granularity Temporal Aware Time Series Classification

引用

IEEE Transactions on Automation science and Engineering 2025年 22卷 16026-16037页

作者： Liang Xi Wei Wang Han Liu Jingqi Pan School of Computer Science and Technology Harbin University of Science and Technology Harbin China

Self-supervised learning (SSL) can extract useful temporal representations for time series classification (TSC) tasks. However, existing methods with subsequence and instance-level augmentation lead to the loss of global information and inductive bias. Meanwhile, neglecting multi-granularity temporal representations also poses challenges for modeling complex temporal structures and relationships. Therefore, in this work, we propose a multi-granularity temporal-aware time series classification method, called MGTC, that learns multi-granularity temporal representations, enhancing the ability to perceive class discrepancies. Specifically, we propose a multi-density masking strategy to adapt the dynamic time-varying patterns for learning comprehensive temporal representations. Next, we employ cluster-wise constraint to hierarchically aggregate these representations at the instance level. Finally, we design two self-supervised tasks: i) granularity-aware contrastive learning, to extract intra-instance fine-grained temporal structures and inter-instance coarse-grained class relationships, and ii) cross-view prediction pretext task, to capture global contextual temporal dependencies. We conducted comprehensive experimental evaluations on various types of dataset, and the experimental results validated the effectiveness of our method. The code is publicly available at https://***/mrxiliang/MGTC Note to Practitioners—Time series classification (TSC) is widely used in many real-world applications, such as industrial fault diagnosis and healthcare monitoring. To guarantee classification accuracy, the TSC models must have high representation learning ability. However, many existing methods fail to capture global information and different granularities of temporal relations, hindering accurate classification. To address this challenge, this paper proposes a multi-granularity temporal-aware time series classification method (MGTC), capturing multi-granularity temporal repr

关键词： Time series analysis Contrastive learning Representation learning Data mining Windows Training Data models Data augmentation Time-frequency analysis Predictive models

来源：评论

学校读者我要写书评

暂无评论

Longitudinal velocity control of autonomous driving based on extended state observer

引用

CAAI Transactions on Intelligence technology 2025年第1期10卷 36-46页

作者： Hongbo Gao Hanqing Yang Xiaoyu Zhang Xiangyun Ren Fenghua Liang Ruidong Yan Qingchao Liu Mingmao Hu Fang Zhang Jiabing Gao Siyu Bao Keqiang Li Deyi Li Danwei Wang Department of Automation University of Science and Technology of ChinaHefeiChina State Key Laboratory of Intelligent Vehicle Safety Technology ChongqingChina Institute of Advanced Technology University of Science and Technology of ChinaHefeiChina School of Electrical and Electronic Engineering Nanyang Technological UniversitySingaporeSingapore School of Traffic and Transportation Beijing Jiaotong UniversityBeijingChina Automotive Engineering Research Institute Jiangsu UniversityZhenjiangChina School of Mechanical and Aerospace Engineering Nanyang Technological UniversitySingaporeSingapore School of Automotive Engineering Hubei University of Automotive TechnologyShiyanChina Beijing Idriverplus Technology Co. Ltd.BeijingChina Chery Automobile Co. Ltd.WuhuChina State Key Laboratory of Automotive Safety and Energy Tsinghua UniversityBeijingChina Department of Computer Science and Technology Tsinghua UniversityBeijingChina

Active Disturbance Rejection Control(ADRC)possesses robust disturbance rejection capabilities,making it well-suited for longitudinal velocity ***,the conventional Extended State Observer(ESO)in ADRC fails to fully exploit feedback from first-order and higher-order estimation errors and tracking error simultaneously,thereby diminishing the control performance of *** address this limitation,an enhanced car-following algorithm utilising ADRC is proposed,which integrates the improved ESO with a feedback *** comparison to the conventional ESO,the enhanced version effectively utilises multi-order estimation and tracking ***,it enhances convergence rates by incorporating feedback from higher-order estimation errors and ensures the estimated value converges to the reference value by utilising tracking error *** improved ESO significantly enhances the disturbance rejection performance of ***,the effectiveness of the proposed algorithm is validated through the Lyapunov approach and experiments.

关键词： active disturbance rejection control(ADRC) autonomous vehicle extended state observer(ESO) intelligent control longitudinal velocity control

来源：评论

学校读者我要写书评

暂无评论

A Real-Time Acceleration-Based Frailty Detection System for Elderly Care 17

A Real-Time Acceleration-Based Frailty Detection System for ...

引用

17th International Convention on Rehabilitation Engineering and Assistive technology, i-CREATe 2024

作者： Sansrimahachai, Watsawee Toachoodee, Manachai Khemapech, Ittipong School of Science and Technology University of the Thai Chamber of Commerce System and Networking Group Computer Science Department Bangkok Thailand University of the Thai Chamber of Commerce System and Networking Group School of Science and Technology Bangkok Thailand

ISBN: (纸本)9798350355154

Over the past decade, almost all countries in the world have experiencing growth in both the size and the proportion of elderly people in the population. The number of world's aging population is expected to reach about 2.1 billion by 2050, which is more than double figure from 1 billion in 2020. This statistic indicates the fact that we are entering the era of 'super-aged society'. One of the major problems in the elderly comes in the form of a decline in health and physical function, generally called Frailty. The elderly with frailty typically suffer several health issues including loss of muscle mass, poor gait and balance and cognitive impairment. They are higher risk of falls and quick progression to dependency. Early identification of frailty is required to effectively reduce and prevent frailty outcomes. To address the problem, we propose a frailty detection system that can remotely analyze gait characteristics and detect frailty among elderly individuals in real-time. Our system utilizes an Ultra-wideband tag attached to the body of the elderly as a data collection device. It applies data stream mining techniques that perform machine learning-based gait analysis and detect frailty based on the patterns of acceleration data. The experimental result indicates that our proposed frailty detection system is practical to be used in real-world settings with relatively high accuracy rate, about 81.8% for overall accuracy. © 2024 IEEE.

关键词： Frailty Detection Gait Analysis Stream Mining

来源：评论

学校读者我要写书评

暂无评论

Federated Learning Privacy-preserving Method Based on Bregman Optimization 10

Federated Learning Privacy-preserving Method Based on Bregma...

引用

10th IEEE International Conference on Cyber Security and Cloud Computing and 9th IEEE International Conference on Edge Computing and Scalable Cloud, CSCloud-EdgeCom 2023

作者： Zhu, Gengming Zhang, Jiyong Zhang, Shaobo Yin, Yijie Hunan University of Science and Technology School of Computer Science and Engineering Xiangtan411201 China

ISBN: (纸本)9798350312461

Federated learning has received a lot of attention for its ability to solve the data silo problem, but it is also limited by the problem of data heterogeneity and privacy. Non-Independent Identical Distribution (Non-I.I.D) data leads to performance degradation of federation models, and privacy problem have been studied as a hot topic in the field of federated learning. However, current research rarely considers non-I.I.D data and privacy simultaneously. In this paper, we propose a federated learning scheme based on Bregman and differential privacy (FLBDP). Our approach adopts Bregman distance for personalized model training, which aims to control the difference between local model and global model in a limited range, the FLBDP can reduce the model difference to improve the model performance by Bregman optimization. In addition, we use a Gaussian mechanism to perturb the personalized model and update the local model by the perturbed personalized model, which enables the model parameters to satisfy differential privacy in the uplink channel to enhance user privacy protection. © 2023 IEEE.

关键词： Learning systems

来源：评论

学校读者我要写书评

暂无评论

FedWM: Federated Crowdsourcing Workforce Management Service for Productive Laziness

FedWM: Federated Crowdsourcing Workforce Management Service ...

引用

2023 IEEE International Conference on Web Services, ICWS 2023

作者： Shi, Zhuan Yao, Zhenyu Yi, Liping Yu, Han Zhang, Lan Li, Xiang-Yang University of Science and Technology of China School of Computer Science and Technology China Univerisity of Liverpool Department of Mathematical Sciences United Kingdom Nankai University College of Computer Science China Nanyang Technological University School of Computer Science and Engineering Singapore

ISBN: (纸本)9798350304855

Federated crowdsourcing, as a dynamic privacy-preserving distributed machine learning approach, has attracted significant research attention recently. Compared to federated learning (FL), clients can dynamically collect and label fresh data as required, and train model on the updated data. Existing research has mainly focused on incentivizing clients to spend more effort on data collection and labelling in order to improve FL model performance. However, as data collection and labeling require human effort, they need to balance work and rest. This need has been overlooked by existing federated crowdsourcing research. In this paper, we propose the Federated Workforce Management (FedWM) approach to bridge this important gap. It first measures the contribution of each client to the FL model, and estimates the urgency collecting new labelled data based on the rate of change of the contribution. Then, FedWM computes the working time taking into account of the client's maximum productivity and self-reported mood. Finally, it takes both the urgency level of obtaining new data and clients' productivity into consideration to provide scheduling services that advise the clients on work-rest balance in a given time slot based on Lyapunov optimization. Through theoretical analysis, we provide the performance bounds of FedWM. Through extensive experiments based on real-world datasets, we demonstrate that FedWM achieves significantly more advantageous tradeoffs between client rest and FL model performance compared to existing approaches. To the best of our knowledge, it is the first federated crowdsourcing framework designed to achieve productive laziness. © 2023 IEEE.

关键词： Crowdsourcing

来源：评论

学校读者我要写书评

暂无评论

Self-Paced Neutral Expression-Disentangled Learning for Facial Expression Recognition 7

Self-Paced Neutral Expression-Disentangled Learning for Faci...

引用

7th Asian Conference on Artificial Intelligence technology, ACAIT 2023

作者： Wu, Zhenqian Li, Xiaoyuan Ren, Yazhou Pu, Xiaorong Zhu, Xiaofeng He, Lifang University of Electronic Science And Technology of China School of Computer Science And Engineering China University of Electronic Science And Technology of China School of Computer Science And Engineering Shenzhen Institute For Advanced Study China Lehigh University Department of Computer Science And Engineering United States

ISBN: (纸本)9798350359145

The accuracy of facial expression recognition is typically influenced by the following factors: high similarities across different expressions, disturbing factors, and the subtle and rapid micro-movements of the face. One potential solution to overcome these obstacles is to leverage the hidden neutral information present in neutral expression images. In this paper, we propose a Self-Paced Neutral Expression-Disentangled Learning (SPNDL) model, which aims to disentangle the neutral information from facial expressions, facilitating the extraction of crucial features and variations. This approach enables the capture of discriminative details within similar expressions and the detection of micro-facial movements. To enhance the learning of these neutral expression-disentangled features (NDFs) and mitigate the challenges posed by non-convex optimization, we propose a self-paced learning (SPL) strategy based on NDFs during the training phase. SPL learns samples from easy to complex by increasing the number of samples selected into the training process, which enables to effectively suppress the negative impacts caused by low-quality samples and inconsistently distributed NDFs. Experiments on three popular databases (i.e., CK+, Oulu-CASIA, and RAF-DB) show the effectiveness of our proposed method. © 2023 IEEE.

关键词： Face recognition

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：