检索结果-内蒙古大学图书馆

6th International Conference on Green Technology and Sustainable Development, GTSD 2022

作者： Pham, Nhat Truong Pham, Van Dung Nguyen-Van, Qui Nguyen, Bao Hung Minh Dang, Duc Ngoc Nguyen, Sy Dzung Institute for Computational Science Ton Duc Thang University Ho Chi Minh City Viet Nam Deep Learning and Application Ho Chi Minh City Viet Nam Ho Chi Minh City University of Technology Department of Telecommunications Engineering Ho Chi Minh City Viet Nam SK-Global JSC Ho Chi Minh City Viet Nam FPT University Computing Fundamental Department Ho Chi Minh City Viet Nam

ISBN: (数字)9781665466288

ISBN: (纸本)9781665466288

Scene text detection and recognition are vital challenging tasks in computer vision, which are to detect and recognize sequences of texts in natural scenes. Recently, researchers have investigated a lot of state-of-The-Art methods to improve the accuracy and efficiency of text detection and recognition. However, there has been little research on text detection and recognition in natural scenes in Vietnam. In this paper, a deep learning-based empirical investigation of Vietnamese scene text detection and recognition is presented. Firstly, four detection models including differentiable binarization network (DBN), pyramid mask text detector (PMTD), pixel aggregation network (PAN), and Fourier contour embedding network (FCEN), are employed to detect text regions from the images. Then, four text recognition models including convolutional recurrent neural network (CRNN), self-Attention text recognition network (SATRN), no-recurrence sequence-To-sequence text recognizer (NRTR), and RobustScanner (RS) are also investigated to recognize the texts. Moreover, data augmentation methods are also applied to enrich data for improving the accuracy and enhancing the performance of scene text detection and recognition. To validate the effectiveness of scene text detection and recognition models, the VinText dataset is employed for evaluation. Empirical results show that PMTD and SATRN achieve the highest scores among the others for text detection and recognition, respectively. For knowledge-sharing, our implementation is publicly available at. © 2022 IEEE.

关键词： optical character recognition

来源：评论

学校读者我要写书评

暂无评论

Double-Machine-Learning-Based Resource Scheduling Method for Offloading Transfers 27

Double-Machine-Learning-Based Resource Scheduling Method for...

引用

27th Asia-Pacific Conference on Communications (APCC) - Creating Innovative Communication Technologies for Post-Pandemic Era

作者： Lin, Xiao Lin, Songlei Li, Yaping Shao, Junyi Shi, Keqin 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)9781665499279

ISBN: (纸本)9781665499279

Dynamically scheduling the bandwidth based on the traffic variation is important for a task offloading system. However, it faces two challenges. On one hand, the time-varying nature of the offloading traffic makes it difficult to be predicted accurately. On the other hand, differentiated mechanisms are applied to different offloading task types, which greatly complicates the behavior of the task offloading system. It is hence difficult to estimate the performance metrics accurately, especially when the metric values are extremely small. To tackle this, we present a double-machine-learning-based resource scheduling (DML-RS) method for task offloading traffic in this paper. The features of DML-RS are as follows: i) the wavelet transform and the sliding time window are incorporated with the LSTM traffic prediction model, which can capture the periodic and volatile natures of the offloading traffic and hence improve the prediction accuracy;ii) the logarithmic converting is applied to the ANN estimation models, which can improve the sensitivity of the ANN models to the small values and hence provides higher estimation accuracy. As a result, DML-RS can predict the traffic demand of the next network reconfiguration time point and optimize the resource allocation based on the performance estimations in advance. Results show that DML-RS offers near-optimal results compared with the existing method.

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

Recloss: Random Epipolar Constraint Loss Function for Supervised optical Flow Estimation models

SSRN

引用

SSRN 2023年

作者： Fan, Zhengyuan Cai, Zemin The Department of Electronic Engineering Shantou University Guangdong Shantou515063 China Key Laboratory of Digital Signal and Image Processing of Guangdong Province Guangdong Shantou515063 China

The majority of supervised models estimate optical flow through minimizing the numerical difference between the predicted flow and the ground truth, resulting in the loss of positional and geometric characteristics of the calculated flow fields. In addition, these models require a large number of parameters and high computational cost when computing optical flow. To address these issues, this paper presents a novel loss function and a lightweight framework for optical flow estimation. The proposed loss function, called the random epipolar constraint loss function (RECLoss), incorporates epipolar geometry into supervised optimization to transform the numerical difference into geometry constraint. The RECLoss can make the optical flow estimation models more effective and enhance their generalization abilities. Moreover, the design of RECLoss is more interpretable and the estimated optical flow fields from RECLoss have clearly defined mathematical meanings. A lightweight recurrent neural network for optical flow estimation (LRFlow) that balances computational cost and estimation accuracy is also proposed. The LRFlow, containing only 3.0M parameters, consists of a feature extractor, a correlation matching module, and an iterative update unit. The proposed lightweight network achieves state-of-the-art results compared to all other lightweight networks on the challenging Sintel and KITTI datasets. The effectiveness of RECLoss and LRFlow is demonstrated through extensive experiments. The feasibility of RECLoss has also been verified by appending to two state-ofthe-art supervised optical flow estimation models, RAFT and GMA. The models trained with RECLoss even outperform the same ones with other improvements. Experimental results show that RECLoss does not entail a substantial increase in computational overhead. Specifically, among the 290K pixel points, a mere 0.9K pixel points are utilized for RECLoss calculation. Remarkably, the models trained with RECLoss exhibit super

关键词： optical flows

来源：评论

学校读者我要写书评

暂无评论

Glaucoma Precognition Based on Confocal Scanning Laser Ophthalmoscopy Images of the Optic Disc Using Convolutional neural network

Glaucoma Precognition Based on Confocal Scanning Laser Ophth...

引用

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Gupta, Krati Goldbaum, Michael Yousefi, Siamak Univ Tennessee Ctr Hlth Sci Dept Ophthalmol Memphis TN 38163 USA Univ Calif San Diego Dept Ophthalmol La Jolla CA 92093 USA Univ Tennessee Ctr Hlth Sci Dept Genet Genom & Informat Memphis TN 38163 USA

ISBN: (纸本)9781665448994

We develop an Artificial Intelligence (AI) framework for glaucoma precognition from baseline confocal scanning laser ophthalmoscopy imaging data, using a convolutional neural network (CNN) model. The proposed framework extracts 'deep features' from convolutional layers of the CNN model, which are used as input to the ensemble learning classifier in order to identify patients that will likely convert to glaucoma after few years. The prediction model achieved area under the receiver operating characteristic curve (AUC) of 0.83 using the data from baseline visit. The model predicted the onset of glaucoma more accurately than known glaucoma risk factors, Glaucoma Probability Score (GPS) and Moorfields Regression Analysis (MRA) parameters of the Heidelberg Retinal Tomograph (HRT) software. The proposed AI construct provides a highly specific and sensitive model that can predict the onset of glaucoma from baseline HRT parameters and has the potential to provide clinicians valuable information regarding the onset of glaucoma.

关键词： optical design optical computing Predictive models optical fiber networks Feature extraction optical imaging optical receivers

来源：评论

学校读者我要写书评

暂无评论

Inference of the optical depth to reionization τ from Planck CMB maps with convolutional neural networks

引用

ASTRONOMY & ASTROPHYSICS 2023年 676卷

作者： Wolz, Kevin Krachmalnicoff, Nicoletta Pagano, Luca Int Sch Adv Studies SISSA Via Bonomea 265 I-34136 Trieste Italy Natl Inst Nucl Phys INFN Sez Trieste Via Valerio 2 I-34127 Trieste Italy Inst Fundamental Phys Universe IFPU Via Beirut 2 I-34151 Trieste Italy Univ Ferrara Dipartimento Fis & Sci Terra Via Saragat 1 I-44122 Ferrara Italy Ist Nazl Fis Nucl Sez Ferrara Via Saragat 1 I-44122 Ferrara Italy Univ Paris Saclay Inst Astrophys Spatiale CNRS F-91405 Orsay France

The optical depth to reionization, & tau;, is the least constrained parameter of the cosmological ? cold dark matter (?CDM) model. To date, its most precise value is inferred from large-scale polarized cosmic microwave background (CMB) power spectra from the High Frequency Instrument (HFI) aboard the Planck satellite. These maps are known to contain significant contamination by residual non-Gaussian systematic effects, which are hard to model analytically. Therefore, robust constraints on & tau;are currently obtained through an empirical cross-spectrum likelihood built from simulations. In this paper, we present a likelihood-free inference of & tau;from polarized Planck HFI maps which, for the first time, is fully based on neural networks (NNs). NNs have the advantage of not requiring an analytical description of the data and can be trained on state-of-the-art simulations, combining the information from multiple channels. By using Gaussian sky simulations and Planck SRoll2 simulations, including CMB, noise, and residual instrumental systematic effects, we trained, tested, and validated NN models considering different setups. We inferred the value of & tau;directly from Stokes Q and U maps at & SIM;4 & DEG;pixel resolution, without computing angular power spectra. On Planck data, we obtained & tau;(NN) = 0.0579 & PLUSMN;0.0082, which is compatible with current EE cross-spectrum results but with a & SIM;30% larger uncertainty, which can be assigned to the inherent nonoptimality of our estimator and to the retraining procedure applied to avoid biases. While this paper does not improve on current cosmological constraints on & tau;, our analysis represents a first robust application of NN-based inference on real data, and highlights its potential as a promising tool for complementary analysis of near-future CMB experiments, also in view of the ongoing challenge to achieve the first detection of primordial gravitational waves.

关键词： cosmic background radiation cosmological parameters dark ages reionization first stars methods data analysis statistical

来源：评论

学校读者我要写书评

暂无评论

Predictions of resonant mode characteristics for terahertz quantum cascade lasers with distributed feedback utilizing machine learning

引用

OPTICS EXPRESS 2021年第10期29卷 15309-15326页

作者： Tang, Ping Chi, Xiaomei Chen, Bo Wu, Chongzhao Shanghai Jiao Tong Univ Ctr Photon & Biophoton Inst Med Robot Shanghai Peoples R China Shanghai Jiao Tong Univ Sch Biomed Engn Shanghai Peoples R China

Terahertz quantum cascade lasers (THz QCLs) are the most powerful solid-state THz sources so far and THz QCLs with various distributed feedback (DFB) gratings have demonstrated single-mode emission, collimated beam, frequency tunability and high output power. Resonant mode characteristics of THz QCLs with DFB, including frequency, loss and electric-field distributions, are important for waveguide analysis, fabrication and indication of THz QCLs' radiative performance. Typically, predictions of these characteristics rely on numerical simulations. However, traditional numerical simulations demand a large amount of running time and computing resources, and have to deal with the trade-off between accuracy and efficiency. In this work, machine learning models are designed to predict resonant mode characteristics of THz QCLs with first-order, second-order, third-order DFB and antenna-feedback waveguides according to the four input structural parameters, i.e. grating period, total length of waveguide, duty cycle of grating and length of highly-doped contact layer. The machine learning models are composed of a multi-layer perceptron for predictions of frequency and loss, and an up-sampling convolutional neural network for predictions of electric-field distribution of the lowest-loss mode, respectively. A detailed study on more than 1000 samples shows high accuracy and efficiency of the proposed models, with Pearson correlation coefficients over 0.99 for predictions of lasing frequency and loss, median peak signal-to-noise ratios over 33.74dB for predictions of electric-field distribution, and the required time of prediction is within several seconds. Moreover, the designed models are widely applicable to various DFB structures for THz QCLs. Resonators with graded photonic heterostructures and novel phase-locked arrays are accurately predicted as examples. (C) 2021 optical Society of America under the terms of the OSA Open Access Publishing Agreement

关键词： Distributed feedback lasers optical feedback Photonic crystal cavities Photonic crystal fibers Quantum cascade semiconductor lasers Resonant modes

来源：评论

学校读者我要写书评

暂无评论

neural network models for Predicting Trends in VTuber Industry on Bilibili

Neural Network Models for Predicting Trends in VTuber Indust...

引用

2022 International Symposium on Robotics, Artificial Intelligence, and Information Engineering, RAIIE 2022

作者： Chen, Jiaqi Department of Mathematics The School of Computing Science Simon Fraser University 8888 University Drive BurnabyBCV5A1S6 Canada

ISBN: (纸本)9781510660137

Due to the paucity of data for this VTuber industry and the volatility of the data itself, in recent years the processing of datasets in the VTuber industry has remained at the statistical stage, and providing accurate forecasts of industry trends has been quite challenging. In this paper, a new way of thinking about dealing with the VTuber industry on Bilibili datasets is conceived, namely using neural networks (NN) and machine learning algorithms to implement predictions. Bilibili is a well-known video-sharing website and also the largest live streaming platform for VTubers in China. More specifically, the prediction of datasets in the VTuber industry using the Long Short-Term Memory (LSTM) model is done based on the TensorFlow machine learning framework. In short, the dataset is processed to match the dimensions of the TensorFlow library input by dividing and preprocessing the training and test datasets, and then the prediction is implemented by the LSTM model. Finally, the error metrics are calculated by Mean Absolute Error (MAE), Mean Squared Error (MSE), and Root Mean Square Error (RMSE). The experimental results show that the LSTM model achieves a very high accuracy match in predicting the VTuber industry dataset, achieves a more competitive performance in data processing, and can provide more effective industry predictions for those engaged in the VTuber industry. © 2022 SPIE.

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

A Heterogeneous CNN Compilation Framework for RISC-V CPU and NPU Integration Based on ONNX-MLIR

引用

IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems 2025年

作者： Cao, Shan Yang, Meiling Liu, Yintao Li, Yu Zhao, Beining Chen, Xinyu Jiang, Zhiyuan Shanghai University Key Laboratory of Specialty Fiber Optics and Optical Access Networks Joint International Research Laboratory of Specialty Fiber Optics and Advanced Communication Shanghai Institute for Advanced Communication and Data Science Shanghai200444 China

With the continuous advancement of convolutional neural networks (CNNs), many neural network processing units (NPU) have emerged in recent years. NPUs offer advantages such as improved energy efficiency and low latency compared to traditional processors. However, NPUs often struggle to keep pace with the growing complexity of algorithmic models, which limits the implementation of certain AI applications. Central processing units (CPU) are known for their versatility, but are computationally inefficient. Combining the strengths of both architectures could present a viable solution to these challenges. Despite this potential, there is currently insufficient academic focus on CPU/NPU co-computing, and few architectures or toolchains have been developed to support such integration. In this paper, we propose a RISC-V CPU/NPU co-computing-based compilation framework to solve the problem of mapping from algorithms to heterogeneous architectures. To bridge the computational gap between heterogeneous platforms, within ONNX-MLIR, we developed three phases to support operator packing, algorithm adaption, and data coordination. To integrate heterogeneous compilation platforms, we introduce four additional functions to handle task scheduling, weight quantization, weight rearrangement, and memory management. In particular: (1) to optimize the transition overhead, an operator packing technique is introduced by extending the ONNX-MLIR framework with custom operators, which enhances NPU efficiency by reducing the number of transitions between operations. (2) At the aspect of task scheduling, a three-mode task scheduler is developed to allow users to customize task distribution for correctness verification, performance optimization, or detailed analysis on heterogeneous platforms, which demonstrates the versatility of our framework. (3) In terms of memory management, an optimized scheme for CPU/NPU architectures is proposed to ensure accurate data access. This memory management emplo

关键词： Information management

来源：评论

学校读者我要写书评

暂无评论

Surrogate neural network model for sensitivity analysis and uncertainty quantification of the mechanical behavior in the optical lens-barrel assembly

引用

COMPUTERS & STRUCTURES 2022年 270卷

作者： Shahane, Shantanu Guleryuz, Erman Abueidda, Diab W. Lee, Allen Liu, Joe Yu, Xin Chiu, Raymond Koric, Seid Aluru, Narayana R. Ferreira, Placid M. Univ Illinois Natl Ctr Supercomp Applicat Urbana IL 61801 USA Univ Illinois Dept Mech Sci & Engn Urbana IL 61801 USA Univ Texas Austin Walker Dept Mech Engn Austin TX 78712 USA Foxconn Interconnect Technol Ltd 2001 S Fifth StSuite 201 Champaign IL 61820 USA

Surrogate neural network-based models have been lately trained and used in a variety of science and engineering applications where the number of evaluations of a target function is limited by execution time. In cell phone camera systems, various errors, such as interferences at the lens-barrel and lens lens interfaces and axial, radial, and tilt misalignments, accumulate and alter profile of the lenses in a stochastic manner which ultimately changes optical focusing properties. Nonlinear finite element analysis of the stochastic mechanical behavior of lenses due to the interference fits is used on highperformance computing (HPC) to generate sufficient training and testing data for subsequent deep learning. Once properly trained and validated, the surrogate neural network model enabled accurate and almost instant evaluations of millions of function evaluations providing the final lens profiles. This computational model, enhanced by artificial intelligence, enabled us to efficiently perform Monte-Carlo analysis for sensitivity and uncertainty quantification of the final lens profile to various interferences. It can be further coupled with an optical analysis to perform ray tracing and analyze the focal properties of the lens module. Moreover, it can provide a valuable tool for optimizing tolerance design and intelligent components matching for many similar press-fit assembly processes. (c) 2022 Elsevier Ltd. All rights reserved.

关键词： Machine learning Finite element analysis Lens assembly Sensitivity analyses Uncertainty quantification High performance computing

来源：评论

学校读者我要写书评

暂无评论

Efficient Service Provisioning in Fog computing for Multimodal Fall Detection Systems

Efficient Service Provisioning in Fog Computing for Multimod...

引用

International Conference on ICT in Business Industry & Government (ICTBIG)

作者： Sarangapani Nivarthi Vijaya N. Aher T. Kuppuraj Ankit Punia Rajiv Gandhi N Lalit Khanna School of Mangement - PG Management Jain (Deemed to be University) Bangalore Karnataka India E&TC Vishwakarma Institute of Technology Pune India Department of Computer Applications Karpagam Academy of Higher Education Coimbatore India Centre of Research Impact and Outcome Chitkara University Rajpura Punjab India Department of Civil Prince Shri Venkateshwara Padmavathy Engineering College Chennai India Chitkara Centre for Research and Development Chitkara University Himachal Pradesh India

ISBN: (数字)9798331518981

ISBN: (纸本)9798331518998

Our capstone project makes use of an actual-world app that can be efficiently managed by an Edge-Fog-Cloud system. Fall detection is an issue that we have taken up and aim to solve by providing decisions with the lowest feasible delay. We propose a three-stage process that makes use of multimodal data (sensors, video, and pictures). An edge device's convolutional neural network infers the visual data used for the first inference. To improve weight distribution and training speed, the neural network is constructed utilizing transfer learning. The optical flow method removes the background noise and focuses just on the individual by representing moving objects as vector fields. A condensed neural network is able to operate on a resource-constrained edge device. After that, the sensor data is classified by the fog layer using SVM, Random Forest (RF) and Gradient Boosting (XGBoost). The fog layer also uses visual and sensor data to make the ultimate fall judgment. The ensemble decision-making process relies on the inferences drawn from the fog and edge layers to arrive at this conclusion. Model construction and durable data storage are carried out at the cloud layer. For more accurate evaluations, our work also deals with false negative rates. To do this, we use augmentation of images on the misclassified photos; this improves the classification and adds value to the dataset. Last but not least, the suggested models are containerized for simple deployment, maintenance, and scalability. With an accuracy of 99%, the proposed technique surpassed all of its competitors, including SVM, XGBoost. Every approach has its advantages, but the Proposed technique clearly outperforms the others and is hence the best option for this specific job.

关键词： Support vector machines Visualization Cloud computing Accuracy Image edge detection neural networks Vectors Sensors Internet of Things Fall 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：