检索结果-内蒙古大学图书馆

International Conference on Computer Vision (ICCV)

作者： Lu Zhang Chenbo Zhang Jiajia Zhao Jihong Guan Shuigeng Zhou Shanghai Key Lab of Intelligent Information Processing and School of Computer Science Fudan University China Science and Technology on Complex System Control and Intelligent Agent Cooperation Laboratory China Department of Computer Science & Technology Tongji University China

Zero-shot object detection aims to localize and recognize objects of unseen classes. Most of existing works face two problems: the low recall of RPN in unseen classes and the confusion of unseen classes with background. In this paper, we present the first method that combines DETR and meta-learning to perform zero-shot object detection, named Meta-ZSDETR, where model training is formalized as an individual episode based meta-learning task. Different from Faster R-CNN based methods that firstly generate class-agnostic proposals, and then classify them with visual-semantic alignment module, Meta-ZSDETR directly predict class-specific boxes with class-specific queries and further filter them with the predicted accuracy from classification head. The model is optimized with meta-contrastive learning, which contains a regression head to generate the coordinates of class-specific boxes, a classification head to predict the accuracy of generated boxes, and a contrastive head that utilizes the proposed contrastive-reconstruction loss to further separate different classes in visual space. We conduct extensive experiments on two benchmark datasets MS COCO and PASCAL VOC. Experimental results show that our method outperforms the existing ZSD methods by a large margin.

关键词：

来源：评论

学校读者我要写书评

暂无评论

An Optimized One-Stage Detector for Autonomous PV Defect Detection

An Optimized One-Stage Detector for Autonomous PV Defect Det...

引用

International Conference on Autonomous Unmanned systems, ICAUS 2021

作者： Huidi, Xiao Jiajia, Zhao Lihui, Wang Qiucheng, Shen Nongjie, Ma Key Laboratory of Micro-Inertial Instrument and Advanced Navigation Technology Ministry of Education School of Instrument Science and Engineering Southeast University Nanjing210096 China Science and Technology on Complex System Control and Intelligent Agent Cooperation Laboratory Beijing100074 China School of Mechanical and Electronic Control Engineering Beijing Jiaotong University Beijing China

ISBN: (纸本)9789811694912

To realize the defect detection of Photovoltaic (PV) modules based on infrared images, a one-stage detector based on FPT and loss function optimization is proposed. Firstly, ResNet50 is selected as the backbone network for feature extraction, combining with the idea of migration learning to strengthen its feature extraction ability. Besides, replacing the traditional feature pyramid with FPT, using Transformer structure help realize the interaction of the same position between different pyramid layers as well as different positions of the same layer, improving the feature fusion capability of the model. Finally, to alleviate the problem of suppression of non-optimal predictions by NMS strategy, a bipartite graph matching algorithm based on loss function minimization is designed to enhance the learning of the distribution of ground truth targets by the network. The dataset is expanded by Mosaic data enhancement method for model training, so as to achieve infrared image-based PV defect detection. Results show that the improved one-stage target detection network has a promising detection effect on PV module defects under infrared images, with the average accuracy improving by 2.20% compared with the original algorithm. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Defects

来源：评论

学校读者我要写书评

暂无评论

A Method for Marine Ship Formation Intention Recognition based on BiConvLSTM-Attention

A Method for Marine Ship Formation Intention Recognition bas...

引用

Electronic Information Engineering and Computer science (EIECS), 2021 International Conference on

作者： Yunqiao Xi Junping Zhang Jian Liu Linlin Wang Hongfeng Xu Linxiu Chen Jixiang Jiang School of Electronics and Information Engineering Harbin Institute of Technology Harbin China Science and Technology on Complex System Control and Intelligent Agent Cooperation Laboratory Institute of Electrical and Mechanical Engineering Beijing China

To solve the problem that current dynamic intention recognition methods fail to make full use of time domain variation information between multi-temporal group targets, which leads to low accuracy of intention recognition. This paper proposes a bidirectional convolutional long short term memory-attention network for marine formation target intention recognition. The network takes the multi-source trajectory data of the marine ship formation target as the input, extracts and uses the target change rule and time domain characteristics of the Marine formation data, and trains the model to have the ability to independently learn the weight of information in different time periods. The simulation results show that the method has good performance and can meet the needs of practical application.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Meta-ZSDETR: Zero-shot DETR with Meta-learning

arXiv

引用

arXiv 2023年

作者： Zhang, Lu Zhang, Chenbo Zhao, Jiajia Guan, Jihong Zhou, Shuigeng Shanghai Key Lab of Intelligent Information Processing School of Computer Science Fudan University China Science and Technology on Complex System Control and Intelligent Agent Cooperation Laboratory China Department of Computer Science & Technology Tongji University China

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Modeling Method of Aircraft Simulation Based on Formal Description of Petri Net 18th

Modeling Method of Aircraft Simulation Based on Formal Des...

引用

18th Chinese intelligent systems Conference, CISC 2022

作者： Zhan, Tianming Huang, Yu Geng, Huapin Tong, Jiahui TAIJI Computer Co. Ltd Beijing100102 China Peking University Beijing100871 China Science and Technology on Complex System Control and Intelligent Agent Cooperation Laboratory Beijing100190 China Beijing Electro-Mechanical Engineering Institute Beijing102299 China

ISBN: (纸本)9789811962028

The simulation of aircraft guidance and control system has characteristics such as multiple model classification, large parameter influence,complex information interaction and so on. The traditional code-level model development method has many limitations, such as time consuming, low reusability, low modeling efficiency, and model quality can not be guaranteed. Based on the concurrency, graphic representation, mathematics and simulation characteristics of Petri net, this paper studies the aircraft simulation modeling method of formal description based on Petri net, constructs a general modeling development and integration software tool that can carry out component, parametric and graphical modeling, solves the problems of multi granularity model modeling and automatic generation of real-time simulation model code, and realizes the visual graphic modeling and comprehensive integration of different granularity models such as components, modules and meta models. The research method has been successfully applied in the simulation modeling of an aircraft. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Petri nets

来源：评论

学校读者我要写书评

暂无评论

SAR Image Registration Based on Feature Points Using Improved CSP-DenseNet

SAR Image Registration Based on Feature Points Using Improve...

引用

Pattern Recognition and Artificial Intelligence (PRAI), International Conference on

作者： Ziwei Wang Yanjiao Yang Guangzhou Qu Qingxi Chi Beijing Electro-Mechanical Engineering Institute Beijing China Beijing Electro-Mechanical Engineering Institute Science and Technology on Complex System Control and Intelligent Agent Cooperation Laboratory Beijing China

Synthetic aperture radar (SAR) image registration is widely used in integrated navigation system with development of guidance system in sensors and other applications. However, there exists some difficulty because of the quality and imaginary principle of SAR images. In this article, a brand-new method based on feature points using improved CSP-DenseNet is proposed to solve the problems of SAR image registration with weak and noisy texture. Deep features of interest points in SAR images are extracted using the proposed network from the crops of search image and template image respectively. The method has an advantage of preserving abundant feature information in SAR images under the influence of insufficient image information and noise of SAR images. The CSP-DenseNet architecture has a cross-stage construction optimizing fused by CSP-Net and DenseNet to extract matching features. The network is partially connected in specific convolutional layers for feature reusing and computation resources saving. Then, Brute Force Matcher and RANSAC voting are successively applied for a larger number of matching pairs and high-precise matching vertexes. Experimental results on various SAR image matching methods show that the proposed method provides better performance than other approaches compared.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Design of Conical Equiangular Spiral Antenna with Wide Bandwidth and High Gain 4

Design of Conical Equiangular Spiral Antenna with Wide Bandw...

引用

4th International Applied Computational Electromagnetics Society Symposium in China, ACES-China 2021

作者： Zong, Liping Zong, Xianzheng Nie, Zaiping Wang, Youcheng Jiang, Peng School of Electronic Science and Engineering UESTC Chengdu China Science and Technology on Complex System Control and Intelligent Agent Cooperation Laboratory Beijing Electro-Mechanical Engineering Institute Beijing China

ISBN: (纸本)9781733509619

In this paper, an ultra-wideband conical equiangular spiral antenna has been designed. The presented antenna covers an operating bandwidth of f-{0}-10f-{0}. In the design, by loading absorbing materials, the voltage standing wave ratio (VSWR) of the antenna in the low frequency is improved, and the addition of metal cone structure increases the gain of high frequency, which makes the antenna performance improved in the whole band. © 2021 Applied Computational Electromagnetics Society (ACES).

关键词： Spiral antennas

来源：评论

学校读者我要写书评

暂无评论

An Optimized Method of Low RCS Carrier Design 4

An Optimized Method of Low RCS Carrier Design

引用

4th International Applied Computational Electromagnetics Society Symposium in China, ACES-China 2021

作者： Zhang, Hanfei Zong, Xianzheng Nie, Zaiping Wang, Youcheng Jiang, Peng School of Electronic Science and Engineering UESTC Chengdu China Science and Technology on Complex System Control and Intelligent Agent Cooperation Laboratory Beijing Electro-Mechanical Engineering Institute Beijing China

ISBN: (纸本)9781733509619

Due to the strong scattering intensity of the apex of the carrier, a scheme of rounding the apex of the carrier is proposed. In this design, it is embodied in face fillet and line fillet. Based on the above designs, this paper compares the smooth transition with the sharp transition carrier, and draws a general conclusion about the low scattering carrier design method. © 2021 Applied Computational Electromagnetics Society (ACES).

关键词： Design

来源：评论

学校读者我要写书评

暂无评论

Lightweight hybrid visual-inertial odometry with closed-form zero velocity update

引用

Chinese Journal of Aeronautics 2020年第12期33卷 3344-3359页

作者： QIU Xiaochen ZHANG Hai FU Wenxing School of Automation Science and Electrical Engineering Beihang UniversityBeijing 100191China Science and Technology on Aircraft Control Laboratory Beijing 100191China Science and Technology on Complex System Control and Intelligent Agent Cooperation Laboratory Beijing 100074China

Visual-Inertial Odometry(VIO) fuses measurements from camera and Inertial Measurement Unit(IMU) to achieve accumulative performance that is better than using individual *** VIO is an extended Kalman filter-based solution which augments features with long tracking length into the state vector of Multi-State Constraint Kalman Filter(MSCKF). In this paper, a novel hybrid VIO is proposed, which focuses on utilizing low-cost sensors while also considering both the computational efficiency and positioning precision. The proposed algorithm introduces several novel contributions. Firstly, by deducing an analytical error transition equation, onedimensional inverse depth parametrization is utilized to parametrize the augmented feature *** modification is shown to significantly improve the computational efficiency and numerical robustness, as a result achieving higher precision. Secondly, for better handling of the static scene,a novel closed-form Zero velocity UPda Te(ZUPT) method is proposed. ZUPT is modeled as a measurement update for the filter rather than forbidding propagation roughly, which has the advantage of correcting the overall state through correlation in the filter covariance matrix. Furthermore, online spatial and temporal calibration is also incorporated. Experiments are conducted on both public dataset and real data. The results demonstrate the effectiveness of the proposed solution by showing that its performance is better than the baseline and the state-of-the-art algorithms in terms of both efficiency and precision. A related software is open-sourced to benefit the community.

关键词： Inverse depth parametrization Kalman filter Online calibration Visual-inertial odometry Zero velocity update

来源：评论

学校读者我要写书评

暂无评论

A Provably Stable SBP-SAT FDTD Subgridding Method without Modifying Yee's Grids

A Provably Stable SBP-SAT FDTD Subgridding Method without Mo...

引用

Antennas and Propagation Society International Symposium

作者： Yuhui Wang Guijie Diao Hong Ni Shunchuan Yang Research Institute for Frontier Science Beihang University Beijing China Beijing Electro-mechanical Engineering Institute Beijing China Science and Technology on Complex System Control and Intelligent Agent Cooperation Laboratory China

A stable finite-difference time-domain (FDTD) sub-grid ding method using the summation-by-parts simultaneous approximation term (SBP-SAT) technique is proposed. To meet the SBP properties and keep Yee's grids unchanged, the projection operators are carefully designed to extrapolate fields on the boundaries. Then, the SATs are used to weakly enforce the boundary conditions between multiple mesh blocks with different mesh sizes. Therefore, its long-term stability is theoretically guaranteed. An iris filter is simulated to validate the effectiveness of the proposed method. Results show that the proposed SBP-SAT FDTD subgridding method is stable, accurate and efficient.

关键词：

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：