检索结果-内蒙古大学图书馆

Magnetic Conference - Short Papers (INTERMAG Short Papers), IEEE International

作者： John Rex Mohan Chisato Yamanaka Ryoyan Feng Arun Jacob Mathew Yoji Nakamura Rohit Medwal Surbhi Gupta Rajdeep Singh Rawat Yasuhiro Fukuma Department of Physics and Information Technology Faculty of Computer Science and Systems Engineering Kyushu Institute of Technology Iizuka Japan Department of Physics Indian Institute of Technology Kanpur Kanpur India Natural Sciences and Science Education National Institute of Education Nanyang Technological University Singapore Research Center for Neuromorphic AI Hardware Kyushu Institute of Technology Kitakyushu Japan

In this work, we employ micromagnetic modelling of a spin Hall oscillator for a direct inference and classification of binary digit inputs. The spectral characteristics of the oscillation is utilized for the classification. We observed a direct inference of binary digit inputs up to a sequence of four binary digits. Subsequently, handwritten digit image recognition is tested with the Modified National institute of Standards and Testing (MNIST) handwritten digit database and acquired an accuracy of 88.6% with a linear classifier network.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Gx2Mol: De Novo Generation of Hit-like Molecules from Gene Expression Profiles via Deep Learning

arXiv

引用

arXiv 2024年

作者： Li, Chen Matsukiyo, Yuki Yamanishi, Yoshihiro Graduate School of Informatics Nagoya University Chikusa Nagoya464-8602 Japan Department of Bioscience and Bioinformatics Faculty of Computer Science and Systems Engineering Kyushu Institute of Technology Fukuoka Iizuka820-8502 Japan

De novo generation of hit-like molecules is a challenging task in the drug discovery process. Most methods in previous studies learn the semantics and syntax of molecular structures by analyzing molecular graphs or simplified molecular input line entry system (SMILES) strings;however, they do not take into account the drug responses of the biological systems consisting of genes and proteins. In this study we propose a deep generative model, Gx2Mol, which utilizes gene expression profiles to generate molecular structures with desirable phenotypes for arbitrary target proteins. In the algorithm, a variational autoencoder is employed as a feature extractor to learn the latent feature distribution of the gene expression profiles. Then, a long short-term memory is leveraged as the chemical generator to produce syntactically valid SMILES strings that satisfy the feature conditions of the gene expression profile extracted by the feature extractor. Experimental results and case studies demonstrate that the proposed Gx2Mol model can produce new molecules with potential bioactivities and drug-like properties. © 2024, CC BY.

关键词： Gene expression

来源：评论

学校读者我要写书评

暂无评论

Single Block Encoder-Decoder Transformer Model for Multi-Step Traffic Flow Forecasting

Single Block Encoder-Decoder Transformer Model for Multi-Ste...

引用

IEEE Symposium on computers & Informatics (ISCI)

作者： Mas Omar Fitri Yakub Mohamad Sofian Abu Talip Mohd Nazmin Maslan Vijay Kumar Sinha Muljono Muljono WEE Laboratory Malaysia-Japan International Institute of Technology Universiti Teknologi Malaysia Kuala Lumpur Malaysia Department of Electronic System Engineering Malaysia-Japan International Institute of Technology Universiti Teknologi Malaysia Kuala Lumpur Malaysia Department of Electrical Engineering Faculty of Engineering University of Malaya Kuala Lumpur Malaysia Faculty of Industrial and Manufacturing Technology and Engineering Universiti Teknikal Malaysia Melaka Malaysia Department of Computer Science & Engineering School of Engineering & Technology Chitkara University Himachal Pradesh India Faculty of Computer Science Universitas Dian Nuswantoro Semarang Indonesia

ISBN: (数字)9798350353853

ISBN: (纸本)9798350353860

Accurate traffic flow forecasting is crucial for managing and planning urban transportation systems. Despite the widespread use of sequence modelling models like Long Short-Term Memory (LSTM) for this purpose, the potential of Transformer models remains underexplored. This is particularly true for the simplest form of a single block encoder-decoder Transformer model, which can be finely tuned through optimised hyperparameters. This paper examines the performance of a singular horizon-step forecasting method for multi-step traffic flow forecasting using a proposed Single Block Encoder-Decoder Transformer model optimised with a Grid Search algorithm. Results demonstrate that this model can enhance forecasting accuracy compared to the state-of-the-art LSTM model typically used for multi-step forecasting. The model effectively captures long-range temporal dependencies within a single road traffic flow dataset. It was tested on hourly traffic flow data to forecast the next 24 hours for the I5-North freeway in California, sourced from the Caltrans Performance Measurement system. The optimal configuration included an embedding dimension of 32, a feed-forward dimension of 128, and 8 attention heads. Results show a significant improvement, with a 4.7% reduction in Root Mean Squared Error compared to an LSTM model with two hidden layers of 100 neurons each, showcasing the potential of Single Block Encoder-Decoder Transformer models for real-world traffic prediction applications.

关键词： Accuracy Neurons Transportation Predictive models Traffic control Transformers Road traffic

来源：评论

学校读者我要写书评

暂无评论

Sensitivity improvement in electro-optic imaging based on polarization CMOS image sensor

Sensitivity improvement in electro-optic imaging based on po...

引用

Asia-Pacific Conference on Microwave

作者： Ryoma Okada Maya Mizuno Hironari Takehara Makito Haruta Hiroyuki Tashiro Jun Ohta Kiyotaka Sasagawa Division of Materials Science Nara Institute of Science and Technology Ikoma Nara Japan National Institute of Information and Communications Technology Koganei Radio Research Institute Tokyo Japan Department of Opto-Electronic System Engineering Chitose Institute of Science and Technology Chitose Hokkaido Japan Faculty of Medical Sciences Kyushu University Maidashi Fukuoka Japan

In this study, we discuss the maximum signal-to-noise ratio (SNR) at 1GHz in the electric field imaging system that we have developed using the electro-optic (EO) effect. To increase the detection sensitivity of electric field imaging, weak polarization changes must be detected. To increase the detection sensitivity while avoiding pixel saturation of the image sensor, we propose a structure in which a uniform polarizer is directly above the polarization image sensor. This increases the polarization modulation and reduces the incident light intensity, enabling a large amount of light to be incident on the EO crystal. The results of SNR measurements by varying the incident light intensity, we consider that an SNR of 57 dB or higher can be expected at an incident light level of 24 dBm or higher.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Observation of out-of-plane spin-orbit torque in a polycrystalline Py/IrMn3 heterostructure

引用

Physical Review B 2024年第24期110卷 245423-245423页

作者： Md Rejaul Karim Sourabh Manna Ayush K. Gupta John Rex Mohan Suman Kumar Maharana Joseph Vimal Vas Arnab Bose Surbhi Gupta Rajdeep Singh Rawat Department of Physics Indian Institute of Technology Kanpur Kanpur 208016 India Natural Sciences and Science Education National Institute of Education Nanyang Technological University 637616 Singapore Department of Physics and Information Technology Faculty of Computer Science and Systems Engineering Kyushu Institute of Technology 680–4 Kawazu Iizuka 820–8502 Japan The Ernst Ruska-Centre for Microscopy and Spectroscopy Forschungszentrum Jülich 52428 Germany Department of Electrical Engineering Indian Institute of Technology Kanpur 208016 India Department of Physics Motilal Nehru National Institute of Technology Barrister Mullah Colony Teliarganj Prayagraj Uttar Pradesh 211004 India

The generation of on-demand out-of-plane spin polarization (σz) promises an efficient field-free switching of perpendicular nanomagnet has been limited mostly in the single crystal materials. For a direct technological application, it is desired to have an out-of-plane spin current generation from polycrystalline sputtered films rather than expensive single crystalline films. Here, we report the observation of out-of-plane and in-plane torques generated by the ẑ and x̂ spin polarization in the polycrystalline antiferromagnet IrMn3/permalloy (Py) heterostructure. A comparatively high value of out-of-plane spin torque ratio of 0.024 has been observed with a large out-of-plane spin Hall conductivity of ≈2.82×104(ℏ2e)(Ωm)−1. Additionally, we have investigated the underline mechanism and the fundamental role of ẑ spins in the context of external field-free magnetization switching of a perpendicular magnetic anisotropy (PMA) material. Our findings provide a new perspective on generation and detection of out-of-plane spin polarization in antiferromagnet material and manipulate the magnetization for the development of the next generation high-density, low-power consumption, logic device applications.

关键词： Spin polarization

来源：评论

学校读者我要写书评

暂无评论

Concurrent Pipeline Rendering Scheme based on GPU multi-queue and partitioning images

Concurrent Pipeline Rendering Scheme based on GPU multi-queu...

引用

2023 International Conference on Optics and Machine Vision, ICOMV 2023

作者： Qiu, Zixia Ma, Junchao Zhang, Haitang Sibahee, Mustafa A.Al. Abduljabbar, Zaid Ameen Nyangaresi, Vincent Omollo College of Big Data and Internet Shenzhen Technology University Shenzhen518118 China College of Applied Technology Shenzhen University Shenzhen518060 China National Engineering Laboratory for Big Data System Computing Technology Shenzhen University Shenzhen518060 China Department of Computer Science College of Education for Pure Sciences University of Basrah Basrah61001 Iraq Technical Computer Engineering Department Al-Kunooze University College Basrah61001 Iraq Huazhong University of Science and Technology Shenzhen Institute Shenzhen518063 China Faculty of Biological & Physical Sciences Tom Mboya University Homabay40300 Kenya

ISBN: (纸本)9781510664777

Compared with rasterization rendering, ray tracing rendering can improve the image’s visual effect and make the image look more realistic. Real-time ray tracing requires very high computing power of graphics processing unit (GPU). When the number of GPUs is limited and the performance of a single GPU cannot be fully utilized, high rendering latency will occur. In this paper, we propose Multi-Queue Concurrent Pipeline Rendering (MQCPR), a novel ray-tracing parallel rendering scheme based on GPU multi-queue. This scheme divides the image area into multiple parts and uses multi-queue of GPU to enable the computation and transmission tasks in the rendering process to be executed simultaneously, which can maximize the performance of a single GPU and improve the graphics rendering speed. MQCPR may keep the GPU busy to make full use of the GPU resources. Experiments illustrate that in the case of a single GPU, compared with the single queue serial rendering scheme, the number of frames per second (FPS) is increased by 1.5 times after using MQCPR. © 2023 SPIE. All rights reserved.

关键词： Graphics processing unit

来源：评论

学校读者我要写书评

暂无评论

Frame Interval Control Technique for High-Frequency Electric Field Imaging by Equivalent-Time Sampling

Frame Interval Control Technique for High-Frequency Electric...

引用

Asia-Pacific Conference on Microwave

作者： Kiyotaka Sasagawa Yoshihiro Akamatsu Ryoma Okada Maya Mizuno Hironari Takehara Makito Haruta Hiroyuki Tashiro Jun Ohta Division of Materials Science Nara Institute of Science and Technology Ikoma Nara Japan Radio Research Institute National Institute of Information and Communications Technology Koganei Tokyo Japan Department of Opto-Electronic System Engineering Chitose Institute of Science and Technology Ikoma Nara Japan Faculty of Medical Sciences Kyushu University Maidashi Fukuoka Japan

We propose an equivalent-time sampling method with controlled frame spacing to improve the intermediate frequency of field imaging systems. Simple equivalent-time sampling with a constant frame interval detects frequency components other than those to be observed. We present a method that cancels frequency components other than the frequency to be observed by lengthening the frame interval by 1/4 cycle every four frames instead of keeping it constant. We show that only the signal of the frequency to be observed can be detected by an image sensor. We also apply this method to electric field imaging.

关键词：

来源：评论

学校读者我要写书评

暂无评论

TDMA-based scheduling for multi-hop wireless sensor networks with 3-egress gateway linear topology

arXiv

引用

arXiv 2021年

作者： Nguyen, Linh Vu Ha, Nguyen Viet Shibata, Masahiro Tsuru, Masato Computer Science and System Engineering Kyushu Institute of Technology Fukuoka Japan

Packet transmission scheduling on multi-hop wireless sensor networks with 3-egress gateway linear topology is studied. Each node generates a data packet in every one cycle period and forwards it bounded for either of gateways at edges. We focus on centrally-managed Time Division Multiple Access (TDMA)-based slot allocations and provide the design of a packet transmission scheduling framework with static time-slot allocation and basic redundant transmission to reduce and recover packet losses. We proposed three general path models, which would represent all possible variations of this topology, and derived a global static time-slot allocation on each of models to maximize the theoretical probability that all packets are successfully delivered to one of the gateways within one cycle period. Copyright © 2021, The Authors. All rights reserved.

关键词： Wireless sensor networks

来源：评论

学校读者我要写书评

暂无评论

The Effect of Data Augmentation on the Accuracy of Fish Species Classification Using Deep Learning

The Effect of Data Augmentation on the Accuracy of Fish Spec...

引用

International Conference on Informatics and Computing (ICIC)

作者： Ellya Helmud Catur Edi Widodo Oky Dwi Nurhayati Doctor Information System Postgraduate Diponegoro University Semarang Indonesia Faculty of Information Technology Information Systems Study Program Atma Luhur Institute of Science and Business Pangkalpinang Indonesia Physics Department Faculty of Science and Mathematics Diponegoro University Semarang Indonesia Department of Computer Engineering Faculty of Engineering Diponegoro University Semarang Indonesia

ISBN: (数字)9798331517601

ISBN: (纸本)9798331517618

Fish image classification presents an intriguing challenge in the field of computer vision. This research aims to develop an accurate classification model to differentiate between four different fish species using a convolutional neural network. The dataset used consists of $\mathbf{3 0 1 0}$ fish images, divided into training, validation, and testing sets. The convolutional neural network model was trained both with and without data augmentation. Evaluation results show that the model trained with data augmentation achieved an accuracy of $95 \%$ with a loss value of 0.0983, slightly better than the model without augmentation which achieved an accuracy of $94.56 \%$ with a loss value of $\mathbf{0. 1 7 9 4}$. This indicates that data augmentation techniques are effective in improving model performance, likely because augmentation helps the model generalize better to variations in fish image data. The results of this research demonstrate the significant potential of convolutional neural network for fish image classification tasks. The developed model can serve as a foundation for the development of computer vision-based applications such as automatic fish species identification in fisheries or educational applications. Further research can be conducted by exploring different convolutional neural network architectures, more advanced data augmentation techniques, and larger datasets to improve model performance.

关键词： Deep learning Training Accuracy Computational modeling computer architecture Fish Data augmentation Data models Convolutional neural networks Image classification

来源：评论

学校读者我要写书评

暂无评论

Deep Learning-Based Pixel Interpolation in Spatial-Domain Phase Shifting Digital Holography

Deep Learning-Based Pixel Interpolation in Spatial-Domain Ph...

引用

Pacific Rim Conference on Lasers and Electro-Optics (CLEO/Pacific Rim)

作者： Shu Kajitani Masanori Takabayashi School of Computer Science and Systems Engineering Kyushu Institute of Technology 680-4 Kawazu Iizuka Fukuoka Japan Faculty of Computer Science and Systems Engineering Kyushu Institute of Technology 680-4 Kawazu Iizuka Fukuoka Japan Research Center for Neuromorphic AI Hardware Kyushu Institute of Technology Kitakyushu Fukuoka Japan

We propose to apply deep learning to interpolate missing pixels in interferograms acquired by the spatial-domain phase-shifting interferometry (PSI), and numerically evaluate the quality of object light reconstructed ...

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：