检索结果-内蒙古大学图书馆

arXiv 2022年

作者： Yang, Maosheng Isufi, Elvin Schaub, Michael T. Leus, Geert Dept. of Intelligent Systems Delft University of Technology Netherlands Dept. of Computer Science RWTH Aachen University Germany Dept. of Microelectronics Delft University of Technology Netherlands

We study linear filters for processing signals supported on abstract topological spaces modeled as simplicial complexes, which may be interpreted as generalizations of graphs that account for nodes, edges, triangular faces, etc. To process such signals, we develop simplicial convolutional filters defined as matrix polynomials of the lower and upper Hodge Laplacians. First, we study the properties of these filters and show that they are linear and shift-invariant, as well as permutation and orientation equivariant. These filters can also be implemented in a distributed fashion with a low computational complexity, as they involve only (multiple rounds of) simplicial shifting between upper and lower adjacent simplices. Second, focusing on edge-flows, we study the frequency responses of these filters and examine how we can use the Hodge-decomposition to delineate gradient, curl and harmonic frequencies. We discuss how these frequencies correspond to the lower- and the upper-adjacent couplings and the kernel of the Hodge Laplacian, respectively, and can be tuned independently by our filter designs. Third, we study different procedures for designing simplicial convolutional filters and discuss their relative advantages. Finally, we corroborate our simplicial filters in several applications: to extract different frequency components of a simplicial signal, to denoise edge flows, and to analyze financial markets and traffic networks. © 2022, CC BY-SA.

关键词： Chebyshev polynomials

来源：评论

学校读者我要写书评

暂无评论

On the Stimulation Artifact Reduction during Electrophysiological Recording of Compound Nerve Action Potentials *

On the Stimulation Artifact Reduction during Electrophysiolo...

引用

Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)

作者： Raphael Panskus Lukas Holzapfel Wouter A. Serdijn Vasiliki Giagka Dept. of Microelectronics Faculty of Electrical Engineering Mathematics and Computer Science Delft University of Technology The Netherlands Dept. of System Integration and Interconnection Technologies Fraunhofer Institute for Reliability and Microintegration IZM Berlin Germany Neuroscience Dept. Erasmus Medical Center Rotterdam the Netherlands

Recording neuronal activity triggered by electrical impulses is a powerful tool in neuroscience research and neural engineering. It is often applied in acute electrophysiological experimental settings to record compound nerve action potentials. However, the elicited neural response is often distorted by electrical stimulus artifacts, complicating subsequent analysis. In this work, we present a model to better understand the effect of the selected amplifier configuration and the location of the ground electrode in a practical electrophysiological nerve setup. Simulation results show that the stimulus artifact can be reduced by more than an order of magnitude if the placement of the ground electrode, its impedance, and the amplifier configuration are optimized. We experimentally demonstrate the effects in three different settings, in-vivo and in-vitro.

关键词：

来源：评论

学校读者我要写书评

暂无评论

3D object detection with low resolution dept. map

3D object detection with low resolution depth map

引用

IEEE International Conference on Consumer Electronics-Taiwan (ICCE-TW)

作者： Wei-Ting Huang Xiu-Zhi Chen Yen-Lin Chen Chieh-Sheng Huang Dept. of Computer Science & Information Engineering National Taipei University of Technology Taipei Taiwan Artificial Intelligence Research and Development Department ELAN Microelectronics Corporation Hsinchu Taiwan

Obtaining high-resolution dept. map from radar and lidar for training dept. estimation network is a common approach in recent years, although the expensive price and high computational complexity cause it hard to be apply in real-world applications. In this research, we propose a novel concept that only requires low resolution dept. map for training, but able to reach similar performance comparing to high-resolution dept. map-based approaches. Refers to the experimental results, which evaluated on KITTI dataset, our proposed concept decreases 7% and 2% on the mAP of car and pedestrian comparing to the original approach in CaDDN, although it achieves similar performance on visualization result.

关键词：

来源：评论

学校读者我要写书评

暂无评论

3D Vehicle Detection for Autonomous Buses

3D Vehicle Detection for Autonomous Buses

引用

IEEE International Conference on Consumer Electronics-Taiwan (ICCE-TW)

作者： Yi-Kai Chiu Xiu-Zhi Chen Yen-Lin Chen Chieh-Sheng Huang Dept. of Computer Science & Information Engineering National Taipei University of Technology Taipei Taiwan Artificial Intelligence Research and Development Department ELAN Microelectronics Corporation Hsinchu Taiwan

The development of automatic driving technology and Advanced Driver Assistance Systems (ADAS) has become more and more mature, and has led to the development of 3D object detection technologies and discussions. Most of the researches are focus on the applications of small cars, but not common in large vehicles, such as buses and trucks, although such applications are more important for them. Therefore, this study mainly uses the bus data to verify the feasibility of the existed model, and by adjusting the loss function, the prediction accuracy of the class and bounding box can be improved. Based on our experimental results, we found that the model trained by KITTI dataset can predict acceptable result refers to the input images from the bus viewpoint. As a result, we can expect that the ADAS for large vehicles can be implemented based on the sample model structure, and its accuracy is able to be improved by fine-tuning based on more large vehicle viewpoint data.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Compensation Method of Quantized Deep Learning Models for Edge Devices

Compensation Method of Quantized Deep Learning Models for Ed...

引用

IEEE International Conference on Consumer Electronics-Taiwan (ICCE-TW)

作者： Xiu-Zhi Chen Jhen-Hao Li Yen-Lin Chen Chieh-Sheng Huang Dept. of Computer Science & Information Engineering National Taipei University of Technology Taipei Taiwan Artificial Intelligence Research and Development Department ELAN Microelectronics Corporation Hsinchu Taiwan

Quantization is one of the optimization methods for developing deep learning models for edge devices. Through converting the floating-point into 8bit integer or even lower bitwidth, the model’s storage size can be reduced. As the rounding error exists during the quantization process, the model performance decreases. As a result, a method that can recover model performance is needed. In this research, a compensation method for improving the performance of quantized deep learning models is proposed, which make the quantized model can achieve equal or even better performance compared to the original floating-point model.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Geometry Details of Inertial Microsensors Influenced on Their Performance

Geometry Details of Inertial Microsensors Influenced on Thei...

引用

2020 Baltic URSI Symposium, URSI 2020

作者： Nazdrowicz, Jacek Stawinski, Adam Napieralski, Andrzej Lodz University of Technology Dept. of Microelectronics and Computer Science Lodz Poland

ISBN: (纸本)9788394942199

In the paper authors take into consideration results of analysis both MEMS accelerometers and gyroscopes under performance that can be changed by some geometry details modifications. Authors consider different types of shapes of springs to show, how they influence the total device operations. These springs are used in both vibratory sensors - accelerometers and gyroscopes. Because principle of operation of both sensors is based on vibratory system response, frequencies play main role in performance assessment. Fundamental quantities which are used for that are natural frequencies and Q-factors. They depend strongly on particular substructures of whole device geometry (because of variation of spring constant and damping coefficients). © 2020 Warsaw University of Technology.

关键词： accelerometer eigenfrequency gyroscope inertial mass MEMS natural frequency spring coefficient suspensions

来源：评论

学校读者我要写书评

暂无评论

Estimation of Comb Structure Capacitance for MEMS Inertial Sensors

Estimation of Comb Structure Capacitance for MEMS Inertial S...

引用

2020 Baltic URSI Symposium, URSI 2020

作者： Nazdrowicz, Jacek Stawinski, Adam Napieralski, Andrzej Lodz University of Technology Dept. of Microelectronics and Computer Science Lodz Poland

ISBN: (数字)9788394942199

ISBN: (纸本)9788394942199

In this paper authors presents very important problem of estimation capacitance structures used in MEMS sensors. The importance comes from fact that during operation, inertial sensors which are built of solid materials deform because of external forces acting on them. Therefore typical capacitor consists of some electrodes changes its value because electrode changes its orientation. Here authors use analytical and FEM results to compare capacitance results and impact on accuracy of measurement. © 2020 Warsaw University of Technology.

关键词： MEMS accelerometer gyroscope capacitance

来源：评论

学校读者我要写书评

暂无评论

Non-monolithic fabrication of thin-film microelectrode arrays on PMUT transducers as a bimodal neuroscientific investigation tool*

Non-monolithic fabrication of thin-film microelectrode array...

引用

Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)

作者： Andrada I. Velea Joshua Wilson Astrid Gollhardt Cyril B. Karuthedath Abhilash S. Thanniyil Vasiliki Giagka Bioelectronics Section Dept. of Microelectronics Faculty of Electrical Engineering Mathematics and Computer Science Delft University of Technology the Netherlands Fraunhofer Institute for Reliability and Microintegration IZM VTT Technical Research Centre of Finland Ltd.

Ultrasound (US)-based neuromodulation has recently emerged as a spatially selective yet non-invasive alternative to conventional electrically-based neural interfaces. However, the fundamental mechanisms of US neuromodulation are not yet clarified. Thus, there is a need for in-vitro bimodal investigation tools that allow us to compare the effect of US versus electrically-induced neural activity in the vicinity of the transducing element. To this end, we propose a MicroElectrode-MicroTransducer Array (MEMTA), where a dense array of electrodes is co-fabricated on top of a similarly dense array of US *** this paper, we test the proof of concept for such co-fabrication using a non-monolithic approach, where, at its most challenging scenario, desired topologies require electrodes to be formed directly on top of fragile piezoelectric micromachined ultrasound transducer (PMUTs) membranes. On top of the PMUTs, a thin-film microelectrode array was developed utilizing microfabrication processes, including metal sputtering, lithography, etching and soft encapsulation. The samples were analysed through focused ion beam–scanning electron microscopy (FIB-SEM), and the results have shown that damage to the membranes does not occur during any of the process steps. This paper proves that the non-monolithic development of a miniaturised bimodal neuroscientific investigation tool can be achieved, thus, opening up a series of possibilities for further understanding and investigation of the nervous system.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Multi-Task Deep Learning Model for Autonomous Driving: Object Detection, Semantic Segmentation, and dept. Estimation

Multi-Task Deep Learning Model for Autonomous Driving: Objec...

引用

IEEE International Conference on Consumer Electronics-Taiwan (ICCE-TW)

作者： Yih-Chen Wang Jhe-Li Lin Yen-Lin Chen Chieh-Sheng Huang Ming-Liang Lai Dept. of Computer Science & Information Engineering National Taipei University of Technology Taipei Taiwan Artificial Intelligence Research and Development Department ELAN Microelectronics Corporation Hsinchu Taiwan Graduate Institute of Intellectual Property National Taipei University of Technology Taipei Taiwan

In the field of autonomous driving, many models based on deep learning methods have been constructed to solve computer vision tasks related to this domain, such as object detection, semantic segmentation, and dept. estimation. Each model can provide different information about the surrounding environment of a vehicle to assist in driving. However, for applying in the real world, more detailed information about the surrounding environment is essentially required. In this study, we propose a model based on the concept of multi-task learning. This model is an encoder-decoder architecture mainly consisting of an encoder using the hard-parameter sharing technique and three decoders for individual tasks. Therefore, this model can handle object detection, semantic segmentation, and dept. estimation at the same time. Our proposed multi-task model has been verified to perform well on the public dataset Cityscapes and has higher generalizability than other models.

关键词：

来源：评论

学校读者我要写书评

暂无评论

RF Performance of FOWLP to PCB Board Transition

RF Performance of FOWLP to PCB Board Transition

引用

Electronics Packaging Technology Conference (EPTC)

作者： K. B. Zheng Sun Mei Sharon Lim Pei Siang Jong Ming Chinq Lau Boon Long Zhou Lin Lim Teck Guan IME attached student from Dept. of Electrical and Computer Engineering National University of Singapore (NUS) Innovis Tower Singapore Institute of Microelectronics (IME) A*STAR (Agency for Science Technology and Research)

ISBN: (纸本)9798350398861

Fan-Out Wafer-Level-Packaging (FOWLP) to PCB transition enables the scaling of an antenna element into an array at a PCB level. This paper presents key design steps ensuring the FOWLP package-to-board transition structure's performance. The back-to-back package-to-board transition structure has a maximum simulated insertion loss of 0.88 dB and a minimum return loss of 11.73 dB over 20 to 35 GHz. A prototype was also fabricated for further characterization.

关键词： Radio frequency 5G mobile communication Wireless networks System performance Web and internet services Prototypes Insertion loss

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：