检索结果-内蒙古大学图书馆

6th International School on Engineering Trustworthy Software systems, SETSS 2024

作者： Liu, Sirui Dong, Wei Ma, Tiecheng Dong, Yanqi Yang, Dong College of Computer Science and Technology National University of Defense Technology Changsha China

ISBN: (纸本)9789819646555

In embedded software, reactive synthesis can automatically generate control automata from Generalized Reactivity of rank 1 (GR(1)) specifications. However, this approach requires handcrafted specifications, and reactive synthesis only provides the automata as the control logic, while the rest of the program must be implemented manually. To address these issues, we propose an automated method for building component-based embedded software through formal synthesis and Large Language Models (LLMs), comprising the following parts: first, we design a component-based embedded software architecture to support our software synthesis method;then, we develop an approach to translate tasks expressed in natural language (NL) into GR(1) specifications based on LLMs, producing the control automaton for our software architecture. We combine the intelligence and simplicity of LLMs with the correctness and safety of formal synthesis. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Embedded software

来源：评论

学校读者我要写书评

暂无评论

Amphi: Practical and Intelligent Data Prefetching for the First-Level Cache

Amphi: Practical and Intelligent Data Prefetching for the Fi...

引用

2025 Design, Automation and Test in Europe Conference, DATE 2025

作者： Tang, Xuan Wang, Zicong He, Shuiyi Dong, Dezun Liao, Xiangke College of Computer Science and Technology National University of Defense Technology Hunan China

ISBN: (纸本)9783982674100

Data prefetchers play a crucial role in alleviating the memory wall by predicting future memory accesses. First-level cache prefetchers can observe all memory instructions but often rely on simpler strategies due to limited resources. While emerging machine learning-based approaches cover more memory access patterns, they typically require higher computational and storage resources and are usually deployed in the last-level cache. Other intelligent solutions for the first-level cache show only modest performance gains. To address this, we propose Amphi, the first practical and intelligent data prefetcher specifically designed for the first-level cache. Applying a binarized temporal convolutional network, Amphi significantly reduces storage overhead while maintaining performance comparable to the SOTA intelligent prefetcher. With a storage overhead of only 3.4 KB, Amphi requires only one-eighth of Pythia's storage needs. Amphi paves the way for the broader adoption of intelligence-driven prefetching solutions. © 2025 EDAA.

关键词： Cache memory

来源：评论

学校读者我要写书评

暂无评论

HyperSDT: HyperNetwork Slide Decision Tree for Interpretable Tabular Learning

HyperSDT: HyperNetwork Slide Decision Tree for Interpretable...

引用

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

作者： Hu, Nan Li, Xueqiong Huang, Jun-Jie Liang, Zhenhua Yang, Shaowu Wang, Ji College of Computer Science and Technology National University of Defense Technology Changsha China

ISBN: (纸本)9798350368741

Recently, substantial progress has been achieved in leveraging deep learning models for tabular data learning. However, despite significant advancements, the predominant focus of these endeavors has been on augmenting the performance of contemporary deep learning models. Consequently, the interpretability of such models is frequently overlooked or rendered secondary, thereby posing a challenge in comprehending their underlying decision-making processes. In this work, we propose a novel HyperNetwork Slide Decision Tree (HyperSDT) approach to achieve interpretable deep learning for tabular data while maintaining a comparable accuracy to state-of-the-art methods. HyperSDT provides a comprehensive interpretable framework with interpretability by using Silde Decision Tree and Decision Transformer together. Our experimental results demonstrate that our framework is competitive with prior baselines under various tabular learning benchmarks while providing better interpretability. The code can be achieved via https://***/hunancreate/HyperSDT. © 2025 IEEE.

关键词： decision tree hypernetwork interpretability tabular data

来源：评论

学校读者我要写书评

暂无评论

Spatio-Temporal Mixed Graph Neural Controlled Differential Equations with Adaptive Connection Sampling for Irregular Multivariate Time Series Anomaly Detection

Spatio-Temporal Mixed Graph Neural Controlled Differential E...

引用

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

作者： Jia, Xudong Peng, Wei Shen, Chiran Zhao, Baokang Xun, Peng College of Computer Science and Technology National University of Defense Technology Changsha China

ISBN: (纸本)9798350368741

Multivariate time series data often demonstrate sparse and irregular characteristics in real-world signal processing applications, making anomaly detection challenging. This paper introduces STMG-AD, a spatio-temporal mixed graph neural controlled differential equation method with adaptive connection sampling, designed specifically for anomaly detection in irregular multivariate time series. By integrating causal graphs with graph attention networks and employing adaptive connection sampling coupled with Monte Carlo dropout, STMG-AD can enhance the robustness and accuracy of anomaly detection. Experiments on various real-world datasets demonstrate its superiority over existing methods in detecting anomalies in irregular multivariate time series. © 2025 IEEE.

关键词： Anomaly Detection Graph Neural Networks Irregular Multivariate Time Series Monte Carlo Dropout Neural Controlled Differential Equations

来源：评论

学校读者我要写书评

暂无评论

A novel commonsense reasoning method based on heterogeneous knowledge fusion and adversarial training

引用

Neural Computing and Applications 2025年第16期37卷 10101-10118页

作者： Du, Yongping Zhang, Qi Yan, Jingya Hou, Ying Han, Honggui College of Computer Science Beijing University of Technology Beijing100124 China Beijing Artificial Intelligence Institute Beijing University of Technology Beijing100124 China

Commonsense question answering (CQA) task is an effective way to measure the ability of language models to understand commonsense knowledge. The previous research focus on combining pre-trained language models (PLMs) with knowledge graphs (KG) for knowledge reasoning. However, these methods still have some limitations, such as the lacking of entity background knowledge in KG, how to fuse heterogeneous information effectively and poor model robustness. To solve the above problems, we propose a novel commonsense reasoning method based on heterogeneous knowledge fusion and adversarial training (HKF-AT). A multi-source knowledge retrieval method is adopted to retrieve entity-related texts in a large corpus and extract the knowledge path from question entity to option entity in KG. Further, multi-source information fusion strategy makes the model utilize and learn rich external knowledge information directly. Specially, the new adversarial samples are generated for both the retrieved texts and triples, and adversarial training is carried out. The experimental results on public benchmark dataset indicate that the proposed method effectively improves the performance and robustness of the model compared with other excellent models. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2025.

关键词： Question answering

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

Federated Prompt Tuning: When is it Necessary? 21st

Federated Prompt Tuning: When is it Necessary?

引用

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

Specialized Accelerator for Arbitrary-Dimension Matrix Operations in Deep Learning 17

Specialized Accelerator for Arbitrary-Dimension Matrix Opera...

引用

17th IEEE International Conference on computer Research and Development, ICCRD 2025

作者： Liu, Fei Liu, Zhouquan Zhang, Jing Lai, Mingche Guo, Hui Huang, Libo National University of Defense Technology College of Computer Science and Technology Changsha China

ISBN: (纸本)9798331531881

This paper addresses the significant challenge of executing inference tasks involving General Matrix Multiplication (GEMM) in deep neural networks(DNN) on resource-constrained edge systems. While previous research has introduced various accelerator architectures for GEMM, these solutions primarily support matrix multiplications of varying dimensions at the software level rather than the hardware level. We propose an optimization of data reuse and memory access patterns through the integration of the outer product method. Building on this, we developed an optimized workload allocation strategy to further minimize data transfer overhead and balance computational resource utilization. Additionally, a heterogeneous hardware module named the Matrix Acceleration Engine (MAE) has been designed and integrated into the RISC-V processor to accelerate prevalent GEMM operations in deep neural networks. Compared to traditional instruction sets, the RV-MAE processor achieves a speedup of up to 194.74× in matrix operations and up to 20.29× in convolutional neural network tasks. Compared to advanced matrix accelerators, the performance is improved by 5.33×. Furthermore, the MAE, evaluated in a 28nm CMOS process, exhibits an area of 9742.18μm2and a power consumption of 2.9mW, thus meeting the low-power demands of edge computing. © 2025 IEEE.

关键词： Deep neural networks

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

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

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：