检索结果-内蒙古大学图书馆

3rd International Conference on Artificial Intelligence and Computer Engineering, ICAICE 2022

作者： Ye, Wen Faculty of Electronic and Information Engineering Xi'an Jiaotong University Xi'an710049 China

ISBN: (纸本)9781510663473

In the process of driving a car, negative emotions such as anger, uneasiness, and nervousness of the driver seriously affect the personal safety of the driver and passengers. However, due to various reasons during the driving process, many factors will prevent the driver from showing true emotions on the face, which will seriously affect the accuracy of identification. And Real-time detection of physiological signals to detect emotions requires complex instruments, which is difficult to carry out in real driving. In order to solve this problem, the authors proposed a framework for driver emotion detection, the proposed algorithm includes three models: (1) Automotive Advanced Driver Assistance System ADAS emotion monitoring framework (2) Deep learning-based human facial expression recognition algorithm convolutional neural network CNN to perform human facial expression recognition. (3) Set a threshold window to decide whether to remind. After 100 epochs of training on the dataset of FER-2013, the model accuracy is 60.5%. The model can accurately classify facial expressions in most cases by volunteer testing. © 2023 SPIE.

关键词： Emotion Recognition

来源：评论

学校读者我要写书评

暂无评论

A CNN-Based optical Music Recognition Method

A CNN-Based Optical Music Recognition Method

引用

Computers, Information Processing and Advanced Education (CIPAE), 2020 International Conference on

作者： Jing Chen Weixiang Gao Kai Zhang School of Automation and Electrical Engineering University of Science and Technology Beijing Beijing China

ISBN: (数字)9798331527662

ISBN: (纸本)9798331527679

optical Music Recognition (OMR) technology is an important means of music digitalization in the future. Today's society is developing in the direction of digitalization, virtualization, and intelligence, and the traditional music field is also deeply affected by it. In this paper, the OMR system mainly includes four parts: the preprocessing of music score images, the localization and deletion of spectral lines, the extraction and recognition of note motifs, and the reconstruction of music information. Finally, Pyqt5 is used to develop a complete system that is convenient for the input of score images, viewing descore images, text information, and playing audio to realize the conversion of scores. The extraction and recognition of note basis uses the direction gradient histogram and Convolutional neural networks (CNN) algorithm to train a model with high accuracy. The local connection, weight sharing, and pooling operation of CNN can effectively reduce the complexity of the network, reduce the number of training parameters, make the model have a certain degree of invariability to translation, distortion, and zoom, and have strong robustness and fault tolerance, and also be easy to train and optimize.

关键词： Training Adaptation models Image recognition neural networks optical distortion optical computing optical fiber networks optical imaging Convolutional neural networks Data mining

来源：评论

学校读者我要写书评

暂无评论

Nonvolatile optical phase shifter-based coherent neural network accelerator for efficient matrix computations and inference tasks

引用

Optics and Laser Technology 2025年 190卷

作者： Zhou, Jiahao Guo, Pengxing Hou, Weigang Sun, Wei Lu, Bing Li, Bin Zhang, Fan Hu, Cong Liu, Weichen Guo, Lei School of Communications and Information Engineering Chongqing University of Posts and Telecommunications Chongqing400065 China Guangxi Key Laboratory of Automatic Detecting Technology and Instruments Guilin University of Electronic Technology Guilin541004 China College of Computing and Data Science Nanyang Technological University 639798 Singapore

This paper introduces a coherent optical neural network (NVCONN) accelerator based on nonvolatile optical phase shifters (NVOPS), designed to perform matrix multiplications and inference tasks with high energy efficiency and precision. Initially, a pixelated programmable NVOPS is designed, assisted by low-loss phase change material Sb2Se3. Each NVOPS can be binary pixelated via pulsed laser direct writing, offering significant reductions in footprint compared to conventional electro-optic and thermo-optic phase shifters. The Mach–Zehnder interferometer (MZI) structure, built with NVOPS, exhibits low insertion loss (30 dB) and arbitrary splitting ratios. By cascading NVOPS-MZIs, a 4 × 4 matrix-scale NVCONN is constructed. Furthermore, the impact of quantization precision on matrix operations for NVCONN is analyzed under four laser spot sizes: full-precision, 0.1, 0.5, and 1.0 μm. The simulation results demonstrate that NVCONN achieves high-precision matrix operations in various quantization situations and shows inference accuracy comparable to 64-bit computers in the ResNet-18 and AugLeNet-5 models. The proposed NVCONN achieves a computing density of 2.84 TOPS/mm2 and an energy efficiency of 1.8 TOPS/W, while maintaining robustness at high data rates (ranging from 10 to 50 Gbps). © 2025 Elsevier Ltd

关键词： Insertion losses

来源：评论

学校读者我要写书评

暂无评论

Fast Cerebral Blood Flow Analysis via Extreme Learning Machine

arXiv

引用

arXiv 2024年

作者： Chen, Xi Zang, Zhenya Li, Xingda Li, David

We introduce a rapid and precise analytical approach for analyzing cerebral blood flow (CBF) using Diffuse Correlation Spectroscopy (DCS) with the application of the Extreme Learning Machine (ELM). Our evaluation of ELM and existing algorithms involves a comprehensive set of metrics. We assess these algorithms using synthetic datasets for both semi-infinite and multi-layer models. The results demonstrate that ELM consistently achieves higher fidelity across various noise levels and optical parameters, showcasing robust generalization ability and outperforming iterative fitting algorithms. Through a comparison with a computationally efficient neural network, ELM attains comparable accuracy with reduced training and inference times. Notably, the absence of a back-propagation process in ELM during training results in significantly faster training speeds compared to existing neural network approaches. This proposed strategy holds promise for edge computing applications with online training capabilities. Copyright © 2024, The Authors. All rights reserved.

关键词： Feedforward neural networks

来源：评论

学校读者我要写书评

暂无评论

Heat transfer solution of casting billet based on physical perception neural network 2

Heat transfer solution of casting billet based on physical p...

引用

2nd International Conference on Artificial Intelligence, Automation, and High-Performance computing, AIAHPC 2022

作者： Zhang, Runlai Northeastern University Shenyang China

ISBN: (纸本)9781510657717

Current analytical methods based on finite volume simulations are inefficient because convection coefficients are often unknown. In this work, a loss function is defined based on the mean error satisfying the PDE, initial conditions, and BC. Furthermore, heat transfer theory is used for feature engineering. The predictions for the one- and two-dimensional cases are verified by comparison with the FE results. The results show that, with engineered features, heat transfer outside the training region can be precisely predicted. Trained models with fixed initial conditions can quickly evaluate a range of BCs and develop feedback loops to implement the Industry 4.0 concept of active manufacturing control based on sensor data. © 2022 SPIE. All rights reserved.

关键词： Heat transfer

来源：评论

学校读者我要写书评

暂无评论

Double-ANN-Aided Resource Partitioning Scheme for Data Transfers in MEC System over EON

Double-ANN-Aided Resource Partitioning Scheme for Data Trans...

引用

IEEE/CIC International Conference on Communications in China (ICCC)

作者： Lin, Xiao Lin, Songlei Li, Yaping Sha, Junyi Li, Yajie Fuzhou Univ Coll Phys & Informat Engn Fuzhou Peoples R China Shanghai Jiao Tong Univ State Key Lab Adv Opt Commun Syst & Networks Shanghai Peoples R China Beijing Univ Posts & Telecommun State Key Lab Informat Photon & Opt Commun Beijing Peoples R China

ISBN: (数字)9781665484800

ISBN: (纸本)9781665484800

In this paper, we use the wavelet transform and the ANN model to predict the dynamic traffic demands. Moreover, we use ANN models to estimate the task performance in an MEC system. By leveraging the ANN predictor and the ANN estimators, we present a resource partitioning scheme for data transfers in the MEC system to dynamically adjust the resource partitioning based on the predicted demands and the performance requirements. Results show that our proposed scheme offers near-optimal results compared with the existing scheme.

关键词： optical networks traffic prediction performance estimation network optimization artificial neural network multi-access edge computing

来源：评论

学校读者我要写书评

暂无评论

Machine Learning at the Edge Using neural network Processor

Machine Learning at the Edge Using Neural Network Processor

引用

IEEE Conference on High Performance Extreme computing (HPEC)

作者： Edwin Lee Michael Parker Michael Cervantes Ben Plotner Raytheon Technology Hanover MD United States Raytheon Technology El Segundo CA United States Raytheon Technology Tewksbury MA United States

Increasingly, power, space, and cooling constrained embedded computing assets need to run machine learning applications to make autonomous, low latency decisions within operational environments. Where GPUs, CPUs, and FPGAs aren't suitable for these constraints, custom ASICs with their non-standard development processes and frameworks are the best available option for bringing the power of machine learning to the edge. This paper discusses a neural network processing architecture and why it surpasses all size, power, throughput, and weight requirements while maintaining performance and allowing interoperability with commercial vendors license comprehensive development environment and machine learning libraries. We compare this neural network processor architecture with existing AI and machine learning platforms that leverage common architectures such as GPUs, CPUs, and FPGAs. The plug and play nature of the neural network processor architecture is optimized for applications such as EO-IR (Electro-optical/Infrared) Sensor and Radar Systems without respinning the hardware, reducing costs. This paper explores best practices for deploying common AI and machine learning models for object detection, super resolution, and natural language processing on a neural network processor at the edge. The neural network processor offer an alternative for deployment of autonomous machine learning at the edge that brings the power of a robust development ecosystem together with an architecture favorable for power and space constrained use cases.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Multi-label Classification of Retinal Diseases using Hybrid Vision Transformer

Multi-label Classification of Retinal Diseases using Hybrid ...

引用

International Conference on computing and networking Technology (ICCNT)

作者： Anshika B.D.K. Patro Department of CSE Rajkiya Engineering College Kannauj

ISBN: (数字)9798350370249

ISBN: (纸本)9798350370270

Retina is the rear side of eye which behaves like a screen for image formation. Retina converts the light energy into the signals for passing these signals further to brain. It is the irreplaceable part of human eye so taking care of retina is the essential task. There are various major retinal diseases like optic disc cupping, age-related macular degeneration, glaucoma, optic disc edema, diabetic retinopathy, etc. which should be detected at earlier stage. Artificial intelligence and machine learning can play a vital role in early detection of retinal diseases and now a days help in improving the quality of care. For detecting these various retinal diseases simultaneously from fundus images, multilabel classification can be used. Multilabel classification predicts each label independently. Firstly, convolutional neural network is used over VIETAI datasets. Transfer learning is also used for training the model in which some pretrained models like DenseNet121, VGG16, ResNet50 are used. These pre-trained models help in enhancing the accuracy. For more improvement in results, vision transformer is used. Then, we used a hybrid model which is combination of convolutional neural network and vision transformer named as hybrid vision transformer. Hybrid vision transformer outperforms of all used algorithms for this multi-label classification problem.

关键词： Integrated optics Training Computer vision Transfer learning Retina Transformers optical imaging Convolutional neural networks Diseases Residual neural networks

来源：评论

学校读者我要写书评

暂无评论

Deep neural networks for Chinese Traditional Landscape Painting Creation 2

Deep Neural Networks for Chinese Traditional Landscape Paint...

引用

2nd International Conference on Artificial Intelligence, Automation, and High-Performance computing, AIAHPC 2022

作者： Yang, Xiaoxi Hu, Jiaxi College of Science and Technology Wenzhou-Kean University Zhejiang Wenzhou325000 China College of Public Administration Sichuan University Sichuan Chengdu610065 China

ISBN: (纸本)9781510657717

Deep learning techniques have been popularly applied for artistic tasks such as turning photographs into paintings or creating paintings in the style of modern art. However, East Asian arts are largely ignored. In this paper, we aim to apply deep learning models to create Chinese traditional landscape paintings. We achieve the goal through two deep learning techniques: image style transfer and image synthesis. We apply the Visual Geometry Group (VGG) network to do the style transfer, which is trained on a pair of a content image and a style image, and the goal is to output an image that renders the target content with the desired style. For the image synthesis, we apply the Deep Convolutional Generative Adversarial network (DCGAN), which requires a large set of painting images to produce as realistic as possible non-exist paintings that mimic the training dataset. © 2022 SPIE. All rights reserved.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Spatiotemporal context-aware network for video salient object detection

引用

neural computing & APPLICATIONS 2022年第19期34卷 16861-16877页

作者： Chen, Tianyou Xiao, Jin Hu, Xiaoguang Zhang, Guofeng Wang, Shaojie Beihang Univ Sch Automat Sci & Elect Engn State Key Lab Virtual Real Technol & Syst Beijing 100191 Peoples R China

It has been witnessed that there is an increasing interest in video salient object detection (VSOD) in computer vision field. Different from image salient object detection (ISOD), VSOD not only requires appearance information but also needs motion cues. Thus, it is essential to exploit spatiotemporal information to generate accurate saliency results. Existing VSOD models mainly combine an ISOD model with long short-term memory (LSTM) or flow-estimation modules to integrate saliency cues estimated from spatial and temporal domain. However, flow-estimation modules heavily rely on optical flow images;the generation process of which is rather time-consuming and severely limits its applications in practice. Besides, the LSTM can only exploit motion cues via a step-by-step propagation in the time domain and is hard to realize the multi-scale spatiotemporal interaction. In this paper, we propose the SCANet to solve the above problems. Specifically, we develop the pyramid dilated 3D convolutional (PD3C) module to generate rich temporal features by leveraging context information. Besides, a feature aggregation module is designed to effectively integrate spatial and temporal features. Equipped with these modules, the SCANet is capable of generating high-quality saliency maps at more than real-time inference speed (41 FPS on a single Titan Xp GPU). Extensive experimental results on six widely used benchmark datasets prove that SCANet outperforms state-of-the-art methods in terms of three standard evaluation metrics. Our code will be publicly available at https://***/clelouch/SCANet.

关键词： Deep learning Multi-level feature integration Spatiotemporal deep features Video salient object detection

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：