检索结果-内蒙古大学图书馆

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

作者： Mei, Jian-Ping Lu, Chunlong Guan, Yuhao Lv, Mingqi College of Computer Science and Technology Zhejiang University of Technology Hangzhou310023 China

ISBN: (纸本)9789819601189

Prompt Tuning (PT) efficiently adapts pre-trained models to downstream tasks. Recent studies have suggested incorporating prompt tuning into the framework of federated learning. This paper aims to answer "whether it is necessary to seek federation when clients already possess strong few-shot learning abilities with local prompt tuning" through experimental studies. We simulated various types of data distribution shifts that may exist among clients in real-world applications and compared the performance of federated PT with its local counterpart. Our results show that the "gain window" of federation in prompt tuning is generally smaller than full-model federated learning, and is directly impacted by the cross-client distribution shift and the amount of data on each client. We have found that implementing federation directly to PT in cross-domain situations can lead to performance degradation, calling for more careful considerations and advanced formulations to ensure optimal results. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

IoT-Supported Indoor Navigation System Using Machine Learning Technique 12th

IoT-Supported Indoor Navigation System Using Machine Learnin...

引用

12th International Conference on Recent Trends in Computing, ICRTC 2024

作者： Christy Jeba Malar, A. Deva Priya, M. Bhuvana, S. Vidhya, K. Sudharsan, S. Suhita, S. Department of Information Technology Sri Krishna College of Technology Tamilnadu Coimbatore India Department of Computer Science and Engineering Karpagam College of Engineering Tamilnadu Coimbatore India School of Computer Science Engineering and Information Systems Vellore Institute of Technology Vellore India

ISBN: (纸本)9789819788354

Indoor Navigation System (INS) supports seamless movement of objects within confined spaces in smart environments. In this paper, a novel INS that relies on ESP32-based Received Signal Strength Indication (RSSI) measurements for precise estimation of distance using K-Nearest Neighbour (K-NN) algorithm is proposed. The system architecture includes three ESP32 nodes functioning as clients, and ESP8266 combined with ESP32 acting as a server. To overcome the limitations of Global Positioning System (GPS) within an indoor environment, Machine Learning (ML)-based prediction models are employed to estimate distance between nodes and predict user locations along with the distance between source and destination. This approach not only addresses the challenges posed by GPS in indoor settings, but also shows the potential of ESP32 devices and predictive algorithms for enhancing navigation accuracy and reliability. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Global positioning system

来源：评论

学校读者我要写书评

暂无评论

Prompt Learning for Few-Shot Question Answering via Self-Context Data Augmentation

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2025年第3期6卷 589-603页

作者： Qiu, Jian-Qiang Zhang, Chun-Yang Chen, C. L. Philip Fuzhou University College of Computer and Data Science Fuzhou350108 China South China University of Technology School of Computer Science and Engineering Guangzhou510006 China

Pretrained language models (PLMs) have shown remarkable performance on question answering (QA) tasks, but they usually require fine-tuning (FT) that depends on a substantial quantity of QA pairs. Therefore, improving the performance of PLMs in scenarios with only a small number of training examples, also known as a few-shot setting, is of great practical significance. Current mitigation strategies for the few-shot QA task largely rely on pretraining a QA task-specific language model from scratch, overlooking the potential of foundational PLMs to generate QA pairs, particularly in the few-shot setting. To address this issue, we propose a prompt-based QA data augmentation method aimed at automating the creation of high-quality QA pairs. It employs the PFT method, adapting the question generation process of PLMs to the few-shot setting. Additionally, we introduce a dynamic text filling training strategy. This strategy simulates the progressive human learning process, thereby alleviating overfitting of PLMs in the few-shot setting and enhancing their reasoning capability to tackle complex questions. Extensive experiments demonstrate that the proposed method outperforms existing approaches across various few-shot configurations. 2691-4581 © 2024 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See https://***/publications/rights/*** for more information.

关键词： Zero-shot learning

来源：评论

学校读者我要写书评

暂无评论

Deep Learning Based Signal Detection for Quadrature Spatial Modulation System

引用

China Communications 2024年第10期21卷 78-85页

作者： Shu Dingyun Peng Yuyang Yue Ming Fawaz AL-Hazemi Mohammad Meraj Mirza School of Computer Science and Engineering Macao University of Science and TechnologyMacaoChina Department of Computer and Network Engineering University of JeddahSaudi Arabia Department of Computer Science College of Computers and Information TechnologyTaif UniversityP.O.Box 11099Taif 21944Saudi Arabia

With the development of communication systems, modulation methods are becoming more and more diverse. Among them, quadrature spatial modulation(QSM) is considered as one method with less capacity and high efficiency. In QSM, the traditional signal detection methods sometimes are unable to meet the actual requirement of low complexity of the system. Therefore, this paper proposes a signal detection scheme for QSM systems using deep learning to solve the complexity problem. Results from the simulations show that the bit error rate performance of the proposed deep learning-based detector is better than that of the zero-forcing(ZF) and minimum mean square error(MMSE) detectors, and similar to the maximum likelihood(ML) detector. Moreover, the proposed method requires less processing time than ZF, MMSE,and ML.

关键词： bit error rate complexity deep learning quadrature spatial modulation

来源：评论

学校读者我要写书评

暂无评论

A Copy-Paste Data Augmentation Method for Urban Tree Detection 21st

A Copy-Paste Data Augmentation Method for Urban Tree Detecti...

引用

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

作者： Fang, Jingyun Dong, Tianyang Fan, Jing College of Computer Science and Technology Zhejiang University of Technology Hangzhou310023 China

ISBN: (纸本)9789819601219

Urban tree detection is crucial for effective city tree management. Recently, deep learning methods have become mainstream for urban tree detection. However, these methods lack sufficient training samples for the diverse types of urban trees and struggle to handle occluded trees. To address these problems, we propose a simple yet effective Copy-Paste data augmentation method for urban tree detection. To address the lack of diverse urban tree training samples, we create an extra urban tree image library and use our Copy-Paste strategy based on this image library. To avoid unrealistic random pasting, we use ViT-Adapter to segment images to obtain the category for each region in the image. In each iteration, we randomly select a point within three categories of regions and paste an image randomly chosen from our image library. To address the problem of poor model performance in complex occlusion scenarios, we introduce more overlapping trees, with an overlap threshold between each pair of pasted trees. Experimenting on the UTD dataset and applying our method to four representative object detection algorithms, we observe a significant improvement in performance. Our method does not increase the parameters and Giga Floating-point Operations Per Second (GFlops) of the original models, and it has almost no impact on frames per second (FPS). © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Query-Efficient Stealing Attacks Against Image Encoders 21st

Query-Efficient Stealing Attacks Against Image Encoders

引用

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

作者： Mei, Jian-Ping Guan, Yuhao Lu, Chunlong Chen, Jie Lv, Mingqi College of Computer Science and Technology Zhejiang University of Technology Hangzhou310023 China

ISBN: (纸本)9789819601240

Encoder stealing attacks pose a new challenge for ensuring the protection of security and safety of those models. Malicious users aim to replicate the target encoder’s functionality by training a stolen encoder with features returned by querying the target encoder. Recent attacks have successfully trained high-performance stolen encoders by aligning predictions between the target and stolen models. However, the cost of queries, a crucial factor in the success of model stealing, has not received adequate attention. In this paper, we propose a query-efficient stealing attack against image encoders. The main idea is to use a small number of query responses as external knowledge to enhance the self-learning ability during stolen model training. We also introduce a learnable perturbator on top of the stolen encoder to perturb the output features of the stolen encoder, further improving its robustness and generalization. Linear evaluation results on image datasets of various settings show that the proposed method achieves comparable or even better performance compared to existing attacks with only 25% of the query cost. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Image coding

来源：评论

学校读者我要写书评

暂无评论

OCT Fingerprint Fusion Combining Cross Attention and Quality Optimization 18th

OCT Fingerprint Fusion Combining Cross Attention and Qualit...

引用

18th Chinese Conference on Biometric Recognition, CCBR 2024

作者： Zhou, Jianru Wang, Haixia Zhang, Yilong Sun, Haohao College of Computer Science and Technology Zhejiang University of Technology Hangzhou310000 China

ISBN: (纸本)9789819610679

Optical Coherence Tomography (OCT) is a non-invasive, high-resolution imaging technology that has recently been used in fingerprint acquisition. The captured external and internal fingerprints are robust to spoofing attacks and resistant to harsh skin conditions. The quality of the OCT fingerprints is limited by imaging mechanism and reconstruction. The paper proposes a fingerprint fusion network leveraging the concept that multiple fingerprints derived from OCT volume data can enhance one another. This network integrates cross-attention and quality optimization techniques to enhance the quality of OCT fingerprint images. Three fingerprints, two internal fingerprints and one external fingerprint reconstructed based on depth or intensity information, are utilized. The network introduces a cross-attention module and a quality index OCL to improve the feature representation capabilities and the retention of high-quality regions. Experimental results show that the fingerprint images obtained by the fusion method in this paper are superior in visual effects and quality scores. abstract environment. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Optical coherence tomography

来源：评论

学校读者我要写书评

暂无评论

DORA: A Reliability-Associated Query Optimization Framework for Plan Selection 25th

DORA: A Reliability-Associated Query Optimization Framework ...

引用

25th International Conference on Web Information systems Engineering, WISE 2024

作者： Zhao, Chaofan Ren, Jing College of Computer Science and Technology National University of Defense Technology Changsha China

ISBN: (纸本)9789819605781

Achieving efficient query optimization is crucial in database management systems (DBMS). In recent years, machine learning models have been widely applied to query optimizers. These models can learn from executed plans to generate and select better query plans. Existing research has demonstrated that learned query optimizers can exhibit significant advantages in relevant query benchmarks after sufficient training. However, the substantial training overhead and unpredictable performance regressions limit their practicality. In this paper, we propose DORA, which employs a DOuble-hints strategy to guide the generation of candidate plans and uses a Reliability-Associated learning-to-rank(LTR) model to select the optimal plan. DORA combines operator hint and leading hint to enable the query optimizer to generate richer candidate plans. It then trains a LTR model to evaluate these candidate plans and clusters them in a fine-grained manner. Each cluster is associated with a reliability value to reflect the accuracy of model’s predictions. We filter out plans in low reliability clusters, thereby eliminating the risk of model regression. Our preliminary experiments show that DORA performs better than the traditional query optimizer in PostgreSQL and can match or even outperform state-of-the-art learned optimizers. Furthermore, DORA can also exhibit excellent optimization performance without fully learning all training data. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Structured Query Language

来源：评论

学校读者我要写书评

暂无评论

Correction to: Adaptive learning-enhanced lightweight network for real-time vehicle density estimation (The Visual computer, (2024), 10.1007/s00371-024-03572-3)

引用

Visual computer 2024年第6期41卷 4393-4394页

作者： Qin, Ling-Xiao Sun, Hong-Mei Duan, Xiao-Meng Che, Cheng-Yue Jia, Rui-Sheng College of Computer Science and Engineering Shandong University of Science and Technology Qingdao266590 China

Correction: The Visual computer https://***/10.1007/s00371-024-03572-3 The publication of this article unfortunately contained mistakes. Figure 3 was not correct. The corrected figure is given below. Wrong figure 3: (Figure presented.) Correct figurure 3: (Figure presented.) The original article has been corrected. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2024.

关键词：

来源：评论

学校读者我要写书评

暂无评论

DACSC: Secure Authentication Protocol Based on Dynamic Authentication Credentials and IntelSGX in Cloud Computing Environments

Journal of Network Intelligence

引用

Journal of Network Intelligence 2023年第3期8卷 948-964页

作者： Ni, Lina Ni, Boguang Tang, Yuncan Zhang, Jinquan College of Computer Science and Engineering Shandong University of Science and Technology Qingdao266590 China

Cloud computing is a new super computing paradigm, which has greatly changed the way users store and process data. Meanwhile, with the prosperity of cloud computing, there are many hidden dangers in data privacy and security. Many secure authentication protocols in cloud computing environment have been proposed. However, most of the authentication schemes are vulnerable to various attacks. Therefore, it is vital to plan a secure and effective authentication protocol in the cloud computing sur-roundings. In this paper, we propose a secure authentication protocol DACSC based on dynamic authentication credential (DAC) and Intel software guard extensions (SGX) for the cloud computing environment. In order to prevent internal attacks and information leakage stored in memory, we store key data in the SGX. Furthermore, we adopt dynamic identity authentication credentials to timely update the server’s pseudo identity information. We conduct a formal safety analysis of our DACSC protocol adopting ProVerif which is a formal security evidence tool under the stochastic prediction model. Moreover, We also conduct other analysis to prove that DACSC can resist internal attacks, simulation attacks, and achieve user anonymity. By comparing security and performance, we concluded that DACSC is relatively secure and protects the privacy of data. © 2023, Taiwan Ubiquitous Information CO LTD. All rights reserved.

关键词： Authentication

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：