检索结果-内蒙古大学图书馆

38th Youth Academic Annual Conference of Chinese Association of Automation, YAC 2023

作者： Fang, Wenxing Su, Yanxu Institutes of Physical Science and Information Technology Anhui University Hefei China School of Artificial Intelligence Anhui University Hefei230601 China Southeast University Key Laboratory of Measurement and Control of Complex System of Engineering Ministry of Education Nanjing210096 China

ISBN: (纸本)9798350303636

With the rapid development of Transformers in the field of computer vision, models based on Transformers have become highly competitive architectures in the area. Although variants of Transformer models have achieved increasing accuracy on image classification tasks, the size of the training set and the number of parameters required by the models have increased dramatically. When dealing with small datasets, such models face problems such as overfitting and undergeneralization, leading to poor accuracy on the test set. We propose a new lightweight vision transformer (LVT) to address these issues. We reconstructed the backbone network, which learns the relationship between pixels through local window self-attention and global self-attention computation. We also use the attention pooling approach to fuse the token sequences generated by the backbone network more meticulously. We trained on the CIFAR-10 and CIFAR-100 datasets from scratch and compared them with a modern convolutional neural network. The experimental results show that LVT outperforms the modern convolutional neural network in terms of accuracy and efficiency. On the test set of CIFAR-10, we have obtained an accuracy of 96.83%, which indicates that our model can effectively solve the problems facing the training of small datasets and has a wide range of application prospects. © 2023 IEEE.

关键词： lightweight pooling self-attention windows

来源：评论

学校读者我要写书评

暂无评论

Language-Queried Target Sound Extraction Without Parallel Training Data

Language-Queried Target Sound Extraction Without Parallel Tr...

引用

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

作者： Ma, Hao Peng, Zhiyuan Li, Xu Li, Yukai Shao, Mingjie Kong, Qiuqiang Liu, Ju School of Information Science and Engineering Shandong University Qingdao China Department of Computer Science North Carolina State University NC United States ARC Lab Tencent PCG China Key Laboratory of System and Control AMSS Chinese Academy of Sciences Beijing China The Chinese University of Hong Kong Hong Kong

ISBN: (纸本)9798350368741

Language-queried target sound extraction (TSE) aims to extract specific sounds from mixtures based on language queries. Traditional fully-supervised training schemes require extensively annotated parallel audio-text data, which are labor-intensive. We introduce a parallel-data-free training scheme, requiring only unlabelled audio clips for TSE model training by utilizing the contrastive language-audio pre-trained model (CLAP). In a vanilla parallel-data-free training stage, target audio is encoded using the pre-trained CLAP audio encoder to form a condition embedding, while during testing, user language queries are encoded by CLAP text encoder as the condition embedding. This vanilla approach assumes perfect alignment between text and audio embeddings, which is unrealistic. Two major challenges arise from training-testing mismatch: the persistent modality gap between text and audio and the risk of overfitting due to the exposure of rich acoustic details in target audio embedding during training. To address this, we propose a retrieval-augmented strategy. Specifically, we create an embedding cache using audio captions generated by a large language model (LLM). During training, target audio embeddings retrieve text embeddings from this cache to use as condition embeddings, ensuring consistent modalities between training and testing and eliminating information leakage. Extensive experiment results show that our retrieval-augmented approach achieves consistent and notable performance improvements over existing state-of-the-art with better generalizability. © 2025 IEEE.

关键词： contrastive language-audio pre-training retrieval-augmented strategy target sound extraction uni-modal training for multi-modal tasks

来源：评论

学校读者我要写书评

暂无评论

Signal Temporal Logic Based Vehicle Trajectory Planning Under Mandatory Stop Rule for Heavy Vehicle Right Turns at Signal-Free Intersections

Signal Temporal Logic Based Vehicle Trajectory Planning Unde...

引用

International Conference on Intelligent Transportation

作者： Wanqing Zuo Yunwen Xu Youren Chen Dewei Li School of Electronic Information and Electrical Engineering Shanghai Jiao Tong University Shanghai China Key Laboratory of System Control and Information Processing Ministry of Education of China Shanghai China

ISBN: (数字)9798331505929

ISBN: (纸本)9798331505936

The blind spots of heavy vehicles during right turns arise a safety concern for traffic. Hence, the regulation mandating heavy vehicles to come to a complete stop before making right turns is crucial in ensuring intersection safety. However, the current trajectory planning methods struggle to integrate these traffic regulations, and the restrictions imposed by such traffic rules greatly impact traffic efficiency at intersections. This paper proposes a signal temporal logic (STL) based trajectory planning method for connected and autonomous vehicles (CAVs) at a signal-free intersection, where traffic rules for heavy vehicles are conveniently described by STL formulas. All CAVs at the intersection are modeled without predefined paths to enhance the flexibility of trajectories. Each STL formula is encoded in a standardized way as a conjunction of inequality and equality constraints for the optimization, and the whole trajectory planning of CAVs is achieved through solving a mixed-integer nonlinear programming (MINLP) problem. Simulation results indicate that our method improves traffic efficiency in the scenario with heavy vehicles right turns.

关键词： Connected vehicles Trajectory planning Simulation Programming Regulation Trajectory Safety Planning Logic Autonomous vehicles

来源：评论

学校读者我要写书评

暂无评论

Thermally Drawn Monolithic Polymer Continuum Robot with Laser Cutting Profile

Thermally Drawn Monolithic Polymer Continuum Robot with Lase...

引用

2023 IEEE International Conference on Robotics and Biomimetics, ROBIO 2023

作者： Liu, Huanghua Zou, Yun Xu, Zheng Gao, Anzhu Shanghai Jiao Tong University Institute of Medical Robotics Department of Automation China Ministry of Education Key Laboratory of System Control and Information Processing Shanghai200240 China Institute of Medical Robotics Shanghai Jiao Tong University Shanghai200240 China

ISBN: (纸本)9798350325706

Continuum robots have the great potential in endoluminal diagnosis and intervention toward narrow and confined workspace. However, fabrication of continuum robots still poses a grant challenge in its miniaturization. To solve this, this paper designs a thermally drawn monolithic polymer continuum robot with laser cutting profile. First, the continuum robot is designed with two degrees of freedom (DOF) for shape deflection, and it consists of four optical fibers with the fiber Bragg grating (FBG) sensor on each for tension sensing as driving cables. Second, its main body is a multi-channel polycarbonate (PC) tube produced by thermal drawing process, and a series of rectangular cutouts are made on it via laser cutting process. By combining these two fabrication processes, the developed continuum robot realizes the integration of the multi-channel structure and the notch feature. Third, a control framework of puller-follower controller using the integrated tension sensing is proposed for teleoperation control. Finally, experimental setup is built for the robot validation, and results show that the repeated positioning accuracy of the robot is 1.08 mm. Moreover, a micro endoscope and a needle are added to the end of the continuum robot, and a phantom study of auripuncture and drug delivery is completed to verify the preliminary feasibility. And the proposed fabrication approach enables the miniaturization and complex structure for the cable-driven continuum robots. © 2023 IEEE.

关键词：

来源：评论

学校读者我要写书评

暂无评论

3D bioprinting of in vitro porous hepatoma models:establishment,evaluation,and anticancer drug testing

引用

Bio-Design and Manufacturing 2024年第2期7卷 137-152页

作者： Xiaoyuan Wang Zixian Liu Qianqian Duan Boye Zhang Yanyan Cao Zhizhong Shen Meng Li Yanfeng Xi Jianming Wang Shengbo Sang Shanxi Key Laboratory of Micro Nano Sensors and Artificial Intelligence Perception College of Electronic Information and Optical EngineeringTaiyuan University of TechnologyTaiyuan 030024China Key Lab of Advanced Transducers and Intelligent Control System of the Ministry of Education Taiyuan University of TechnologyTaiyuan 030024China Shanxi Institute of 6D Artificial Intelligence Biomedical Science Taiyuan 030031China Shanxi-Zheda Institute of Advanced Materials and Chemical Engineering Taiyuan 030024China Department of Pathology Shanxi Provincial Cancer HospitalTaiyuan 030024China General Hospital of Taiyuan Iron and Steel Company Taiyuan 030809China

Traditional tumor models do not tend to accurately simulate tumor growth in vitro or enable personalized treatment and are particularly unable to discover more beneficial targeted *** address this,this study describes the use of threedimensional(3D)bioprinting technology to construct a 3D model with human hepatocarcinoma SMMC-7721 cells(3DP-7721)by combining gelatin methacrylate(GelMA)and poly(ethylene oxide)(PEO)as two immiscible aqueous phases to form a bioink and innovatively applying fluorescent carbon quantum dots for long-term tracking of *** GelMA(10%,mass fraction)and PEO(1.6%,mass fraction)hydrogel with 3:1 volume ratio offered distinct pore-forming characteristics,satisfactorymechanical properties,and biocompatibility for the creation of the 3DP-7721 *** analysis and quantitative real-time fluorescence polymerase chain reaction(PCR)were used to evaluate the biological properties of the *** with the two-dimensional culture cell model(2D-7721)and the 3D mixed culture cell model(3DM-7721),3DP-7721 significantly improved the proliferation of cells and expression of tumor-related proteins and ***,we evaluated the differences between the three culture models and the effectiveness of antitumor drugs in the three models and discovered that the efficacy of antitumor drugs varied because of significant differences in resistance proteins and genes between the three *** addition,the comparison of tumor formation in the three models found that the cells cultured by the 3DP-7721 model had strong tumorigenicity in nude *** evaluation of the levels of biochemical indicators related to the formation of solid tumors showed that the 3DP-7721 model group exhibited pathological characteristics of malignant tumors,the generated solid tumors were similar to actual tumors,and the deterioration was *** research therefore acts as a foundation for the application of 3DP-7721 models in drug devel

关键词： 3D bioprinting Hepatoma tumor models Drug screening Antitumor drug development

来源：评论

学校读者我要写书评

暂无评论

User-assisted Caching Strategy Based on Reverse Auction and Recommendation in Railway Station 5

User-assisted Caching Strategy Based on Reverse Auction and ...

引用

5th IEEE International Conference on Electronics Technology, ICET 2022

作者： Sun, Chenghao Wu, Hao Beijing Jiaotong University State Key Laboratory of Rail Traffic Control and Safety School of Electronics and Information Engineering Beijing China Beijing Jiaotong University State Key Laboratory of Rail Traffic Control and Safety Frontiers Science Center for Smart High-speed Railway System Beijing China

ISBN: (纸本)9781665485081

In the field of high-speed railway communication, solving the growing demand for multimedia services of passengers has always been a problem. To provide users with stable video services, caching technology has received more and more attention. In this paper, we refine the details when users assist the on-board relay to cache videos in railway stations. Based on this scenario, we improve the existing asynchronous allocation strategy and propose a user-assisted caching strategy named sequential allocation algorithm with recommendation for caching videos. The experimental results show that the algorithm is superior to synchronous and asynchronous allocation algorithm in social revenue and user participation ratio. © 2022 IEEE.

关键词： Multimedia services

来源：评论

学校读者我要写书评

暂无评论

Thermal analysis and package optimization for the multi-wavelength laser bar module

Thermal analysis and package optimization for the multi-wave...

引用

2022 International Symposium on Semiconductor and Electronic Technology, ISSET 2022

作者： Sun, Wanjun Yang, Lijiao Xie, Jianan Zhang, Xiaxia Sha, Yang Mi, Shuai Lin, Tao College of Automation and Information Engineering Xi’an University of Technology Shaanxi Xi’an710048 China Shaanxi Key Laboratory of Complex System Control and Intelligent Information Processing Xi’an University of Technology Shaanxi Xi’an710048 China

To meet the demand of the laser application market for semiconductor laser function diversification, this paper proposes a handheld multi-wavelength laser stack packaging module. The 808 nm, 915 nm, and 980 nm semiconductor laser bars are selected for the integrated assembly of the multi-wavelength module. However, with the continuous increase of the optical power density of the integrated packaging module, thermal management has become a key factor affecting the performance and life of the device. In this paper, the finite element analysis method is used to analyze the influence of the arrangement of three laser stripe chips in the stacked packaging structure on the heat dissipation of the multi-wavelength module. The simulation results show that the heat dissipation efficiency of the multi-wavelength module can be improved by 13.74 % by changing the arrangement of chips. A back heat dissipation packaging structure with higher heat dissipation efficiency is proposed, and the influence of chip arrangement and working mode on the thermal characteristics of the device in the new packaging structure is simulated and analyzed. According to the simulation results, a chip arrangement that can make the heat dissipation effect of the multi-wavelength module achieve the best is determined. © 2022 Institute of Physics Publishing. All rights reserved.

关键词： Semiconductor lasers

来源：评论

学校读者我要写书评

暂无评论

TL-KDNET: A Malicious Traffic Detection Model Integrating Transfer Learning and Knowledge Distillation 21

TL-KDNET: A Malicious Traffic Detection Model Integrating Tr...

引用

21st International Computer Conference on Wavelet Active Media Technology and information processing, ICCWAMTIP 2024

作者： Yu, Kun Huang, YuanYuan Zhang, Linshan Lv, Jun Lu, Jiazhong School of Cybersecurity Xin Gu Industrial College Chengdu University of Information Technology Chengdu610225 China Advanced Cryptography and System Security Key Laboratory of Sichuan Province Chengdu610225 China SUGON Industrial Control and Security Center Chengdu610225 China Co. Ltd. China

ISBN: (纸本)9798331519254

With the evolution of cyberattacks, traditional methods for malicious traffic detection face increasingly complex threats. To address this challenge, we propose a novel model, TL-KDNet, which integrates transfer learning, domain adaptation, and knowledge distillation techniques to enhance the accuracy and generalization capabilities of malicious traffic detection. In our model, KullbackLeibler (KL) divergence serves as a core module to quantify the differences in probability distributions between the source and target domains, thereby effectively facilitating knowledge transfer from known source domains to unknown target domains. Additionally, by leveraging Maximum Mean Discrepancy (MMD) and a single-sample distillation loss function, we achieve alignment of domain features and transfer of knowledge. Experimental results demonstrate that TL-KDNet improves detection accuracy by 4% on public datasets compared to existing methods. © 2024 IEEE.

关键词： Domain Knowledge

来源：评论

学校读者我要写书评

暂无评论

LOGNET: A Knowledge Distillation-Based Model for Log Anomaly Detection 21

LOGNET: A Knowledge Distillation-Based Model for Log Anomaly...

引用

21st International Computer Conference on Wavelet Active Media Technology and information processing, ICCWAMTIP 2024

作者： Yu, Kun Huang, Yuanyuan Zhang, Linshan Lv, Jun Lu, Jiazhong School of Cybersecurity Xin Gu Industrial College Chengdu University of Information Technology Chengdu610225 China Advanced Cryptography and System Security Key Laboratory of Sichuan Province Chengdu610225 China SUGON Industrial Control and Security Center Chengdu610225 China Co. Ltd. China

ISBN: (纸本)9798331519254

Due to the diverse nature of attack forms, supervised log anomaly detection models require a large volume of labeled data, making the annotation process time-consuming and labor-intensive. Conversely, unsupervised log anomaly detection models often struggle to effectively capture log features and accurately identify anomalous logs. To address these challenges, this study constructs a collaborative learning framework and proposes the LogNet model. This framework enables mutual learning between two distinct models, enhancing their performance synergistically. In the LogNet model, log data is first converted into directed graphs, leveraging the strengths of Graph Convolutional Networks (GCN) and Graph Attention Networks (GAT). Through collaborative learning, the two models guide each other, boosting their detection capabilities. The effectiveness of LogNet is validated on five public datasets, with experimental results demonstrating that LogNet significantly outperforms traditional models in malicious log detection, achieving up to a 4 % improvement in accuracy. © 2024 IEEE.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Self-Supervised Learning with Min-Max Distance Approach for Few-Shot Steel Surface Defects Recognition 43

Self-Supervised Learning with Min-Max Distance Approach for ...

引用

43rd Chinese control Conference, CCC 2024

作者： Wang, Tianlei Chen, Xinyao Li, Zeliang Zhou, Zhencai Zhou, Zhiheng Wuyi University School of Electronic and Information Engineering Jiangmen China Jiangmen Key Laboratory of Kejie Semiconductor Bonding Technology and Control System Jiangmen China Wuyi University School of Mechanical and Automation Engineering Jiangmen China Guangdong Province Kejie Machinery Automation Company Technical Department Jiangmen China

ISBN: (纸本)9789887581581

Surface defects in steel are crucial for quality inspection in the manufacturing of industrial metal products. While deep learning methods have been extensively studied in this field, the lack of large-scale annotated defect samples imposes significant labeling pressure, making training fully supervised deep models challenging. In response to this issue, we employ self-supervised learning for few-shot recognition, effectively learning representations from limited unlabeled data and achieving defect recognition with a small amount of labeled samples. In particular, we construct Other encoder with multi-view information and Base encoder with single view information based on the distillation self-supervised learning framework through the multi-crops mechanism, and the Other encoders and Base encoders perform information share. Our main contribution is that, based on this framework, we propose a new regularization method to solve the problem of instances deviating from the center of self-supervised clustering by minimizing the maximum instance distance (MMD). In testing, fine-tuning the MMD pre-trained model with a small number of labeled data achieves superior recognition performance on the steel surface defect dataset NEU, significantly outperforming other testing methods. © 2024 Technical Committee on control Theory, Chinese Association of Automation.

关键词： Self-supervised learning

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：