检索结果-内蒙古大学图书馆

IEEE International Symposium on Quality Electronic Design

作者： Dantu Nandini Devi Gandi Ajay Kumar Bindu G Gowda Madhav Rao IIIT-Bangalore Bangalore India

ISBN: (数字)9798350309270

ISBN: (纸本)9798350309287

The utilization of hardware-designed approximate computing in Convolutional Neural Networks (CNNs) offers notable advantages, including accelerated performance, enhanced power efficiency, and a compact design footprint. Systolic Array (SA) architectures, optimized for matrix multiplication and convolution operations, have been extensively studied in the context of stand-alone image processing applications. However, their potential for CNN workloads has not been thoroughly assessed. SAs consist of an array of Processing Elements (PEs) structured to perform product operations and accumulations. Incorporating inexact computing units in the SA introduces deviations from precise results, posing a challenge for sustaining hardware accelerator designs in CNN workloads. This paper presents a strategy for the optimal placement of both positive and negative error-distributed multipliers as PE elements to create an error-diluted SA structure. The proposed strategy to structure SA is evaluated for prewitt filter and three other filters extracted from first layer of *** paper introduces an optimization framework for selecting the most suitable PEs from a pool of positive and negative error-distributed multipliers, aiming to achieve a balance between hardware efficiency and image quality metrics. Furthermore, the framework and hardware design files are made available for further usage to the designers’ and researchers community.

关键词： Costs Computer architecture Parallel processing filtering algorithms Hardware Systolic arrays Delays

来源：评论

学校读者我要写书评

暂无评论

Recurrent Neural Network Data Predistortion for Transparent Satellite Channels

Recurrent Neural Network Data Predistortion for Transparent ...

引用

IEEE Conference on Global Communications (GLOBECOM)

作者： David Kopyto Gerhard Bauch Institute of Communications Hamburg University of Technology Hamburg Germany

ISBN: (数字)9798350351255

ISBN: (纸本)9798350351262

Satellite transponders operate with traveling wave-tube amplifiers (TWTAs) which nonlinearly distort the transmit signal, particularly when transmission power close to amplifier saturation is desired. Hence, the concatenation of pulse shaping filter, input multiplexer (IMUX) filter, TWTA, output multiplexer (OMUX) filter and receive filter causes distorted decision regions and severe intersymbol interference (ISI) at the receive filter output. The standards DVB-S2 and DVB-S2X recommend setups for transparent satellite channels. Here, the distortion is traditionally compensated using a lookup table-based centroid data predistortion algorithm. In this work, we propose an improved approach for data predistortion based on bidirectional recurrent neural networks (BRNNs). We show the superiority of BRNN data predistortion over centroid-based predistortion by comparing information rates and bit error rates on recommended setups from the DVB-S2 standard. Furthermore, we show how BRNN predistortion enables ISI compensation for high order modulations from DVB-S2X, where the traditional lookup table approach fails due to infeasible memory requirements.

关键词： Multiplexing Satellites Recurrent neural networks Bit error rate Symbols filtering algorithms Distortion Predistortion Standards Information rates

来源：评论

学校读者我要写书评

暂无评论

An Iterative Method to Effectively Filter-Out Harmonics Injected by High-Speed Railways

An Iterative Method to Effectively Filter-Out Harmonics Inje...

引用

National Power Systems Conference (NPSC)

作者： Rishika Jangde Archita Vijayvargia Anandarup Das Abhijit R. Abhyankar Department of Electrical Engineering Indian Institute of Technology Delhi New Delhi India

ISBN: (数字)9798331519568

ISBN: (纸本)9798331519575

High-speed railway (HSR) operations are increasingly recognized for their sustainability in electric transportation. However, the nonlinear characteristics of traction loads introduce harmonic currents into power systems. These harmonics generated during HSR operations have a critical frequency range of 1.5-2 kHz, which significantly impact the grid by increasing system losses and degrading equipment performance. Addressing this issue requires effectively filtering out these harmonics at the interface of traction systems to prevent distortions from entering the power supply. Existing filter designs for traction often overlook higher-order frequencies, and there is limited focus on the optimal placement and use of filters for reactive power compensation. This paper proposes an effective harmonics filtering technique to regulate the impact of harmonics generated by onboard HSRs within permissible limits. Additionally, a heuristic algorithm is developed to determine the optimal placement of these filters along the feeding section and analyze the reactive power compensation provided by the filters. The effectiveness of this approach is demonstrated on a real HSR track in India. The results validated against the IEEE-519 standard for harmonic control in electrical power systems, show significant improvements in total harmonic distortions of current and voltage compared to instances without filters.

关键词： Reactive power Total harmonic distortion Passive filters filtering filtering algorithms Harmonic analysis Active filters Distortion Rail transportation Power harmonic filters

来源：评论

学校读者我要写书评

暂无评论

Research on Toxic Speech Detection Based on Large Language Models

Research on Toxic Speech Detection Based on Large Language M...

引用

IEEE International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom)

作者： Weihao Li Yongbing Gao Yu Zhang Lidong Yang Ruiping Gao Digital Industry College Inner Mongolia University of Science & Technology Baotou China Technical Network Team Baotou Public Security Bureau Baotou China

ISBN: (数字)9798331506209

ISBN: (纸本)9798331506216

Toxic Speech (TS) is widely spread in social media and is increasingly harmful to the online networking environment. At the same time, words and their variants containing implicit toxic semantics are influenced by the humanistic and social environment of the language and are extremely rich in form and shape, and the lack of background knowledge related to the "discourse body" of existing research has led to the confusion of Implicit Toxic Speech (ITS) with Non-Toxic Speech (NTS). Therefore, this paper proposes the Prompt-enhanced Attention Head Fusion (PAHF) toxic text classification model based on the generative language model. We optimize the hidden layer vectors by intervening in the hidden layer information through prompt, then filter the vectors by combining linear probes to locate those vectors in the hidden layer information that are strongly correlated with the encoding of the background knowledge, and finally construct an attention fusion mechanism to fuse the filtered features, which provides a higher detection accuracy compared with the existing algorithms.

关键词： Maximum likelihood detection Social networking (online) Text categorization Semantics Speech recognition Nonlinear filters filtering algorithms Information filters Vectors Probes

来源：评论

学校读者我要写书评

暂无评论

Spam Email Detection Using HAN-LSTM Network Optimized with Novel Hybrid Bonobo Optimizer and Hunger Games Search Algorithm

Spam Email Detection Using HAN-LSTM Network Optimized with N...

引用

Data Science and Network Security (ICDSNS), International Conference on

作者： Indu Singh Hemanth Siddharth Desugari Shaurya Wadehra Venkata Sai Karthik Department of Computer Science and Engineering Delhi Technological University New Delhi India

ISBN: (数字)9798350373110

ISBN: (纸本)9798350373127

In today's digital world, spam email has grown to be a serious problem that causes problems for email service providers as well as subscribers. Creating a reliable system that can effectively filter out spam emails is required to address this problem. This paper proposes a novel approach to spam email detection that combines advanced feature engineering and hybrid metaheuristic optimization. The proposed method employs Bidirectional Encoder Representations from Transformers (BERT) to derive dense feature vectors, followed by a HAN-LSTM network for improved feature representation and classification. Additionally, we introduce a novel hybrid metaheuristic optimization algorithm comprising of Bonobo Optimizer and Hunger Games Search (hBOHGS) that combines the collaborative exploration of Bonobo Optimizer with the competitive selection of Hunger Games Search to efficiently find optimal values for the parameters in the HAN-LSTM network. Our experimental evaluation on the Enron 1 spam dataset demonstrates the effectiveness of our methodology, achieving an accuracy of 98.04%, a recall of 0.99, and a precision of 0.98. These findings highlight the practical applicability of our system in effectively detecting spam emails, providing a promising alternative in the field of spam email detection.

关键词： Unsolicited e-mail Games Network security filtering algorithms Data science Transformers Feature extraction Vectors Encoding Reliability

来源：评论

学校读者我要写书评

暂无评论

Particle Estimation Algorithm Using Angle between Observation Vectors for Nonlinear System State

引用

Chinese Journal of Electronics 2023年第4期18卷 700-702页

作者： Jun Liang Yu Peng Xiyuan Peng Department of Automatic Test and Control Harbin Institute of Technology Harbin China

A particle estimation algorithm where the weight of the particle is related to angle between observation vectors is presented for nonliear system state. When the likelihood has a bimodal nature, this algorithm leads to more accurate state estimates than Sequential importance resampling (SIR), Auxiliary particle filter (APF), Regularized particle filter (RPF), and Gaussian particle filter (GPF).

关键词： filtering algorithms Particle filters State estimation Nonlinear systems

来源：评论

学校读者我要写书评

暂无评论

Average Weight Optimization RBPF Method for Target Tracking in Multi-Sensor Observations

引用

Chinese Journal of Electronics 2023年第2期22卷 401-404页

作者： Xianxing Liu Zhentao Hu Jie Li Institute of Image Processing and Pattern Recognition Henan University Kaifeng China

The reasonable design of particle filter framework in multi-sensor observation system is the key to expand the application domain of sampling nonlinear filters. Aiming at the effective realization of particle filter for multi-sensor target tracking problem, a novel average weight optimization Rao-Blackwellised particle filtering algorithm is proposed. Combining with the kinetic equation of target state evolution, RBPF is used as the basic estimator of algorithm realization. For the rational utilization from multi-sensor observations and the reduction of the adverse influence from random observations noise in measuring process of particles weight, the average weight optimization strategy is used to improve the reliability and stability of particle weight variance. In addition, we give the concrete flow of RBPF in average weight optimization strategy. Finally, the theoretical analysis and experimental results show the feasibility and efficiency of the proposed algorithm.

关键词： Weight measurement Technological innovation Target tracking Redundancy filtering algorithms Reliability theory Particle filters

来源：评论

学校读者我要写书评

暂无评论

Switching filtering Algorithm with Image Preprocessing and Gesture Tracking

Switching Filtering Algorithm with Image Preprocessing and G...

引用

Chinese Control and Decision Conference, CCDC

作者： Zesheng Hu Lei Yu Shumin Fei School of Mechanical and Electrical Engineering Soochow University Suzhou School of Automation Southeast University Nanjing

ISBN: (纸本)9781665478977

Denoising has always been a critical technology in the image processing community. Gaussian noise and impulse noise are the most common image noises. A single filtering algorithm cannot suppress these two noise conditions. Besides, existing hybrid flitering methods have high time complexity and are difficult to meet the real-time requirements. For this, we propose a filtering algorithm based on conditional switching, which establishes an adaptive median filtering subsystem and a fast guided filtering subsystem. It can automatically determine the main noise and select a suitable filtering subsystem for image denoising. To prove the effectiveness of our method, the proposed method is apply to a Kinect-based gesture tracking system. Experimental results show that the switching filtering algorithm can fulfill both accuracy and real-time performance requirements.

关键词： Location awareness filtering Noise reduction Switches filtering algorithms Real-time systems Noise measurement

来源：评论

学校读者我要写书评

暂无评论

Performance evaluation of a many/multi-core implementation for signal filtering problem

Performance evaluation of a many/multi-core implementation f...

引用

International IEEE Conference on Signal-Image Technologies and Internet-Based System

作者： Vincenzo D’Alò Pasquale De Luca Science and Technologies Department Parthenope University of Naples Naples Italy

Signal filtering plays an important role in preprocessing procedures used in several application fields. In last years, several methods have been proposed by achieving good results. When large data are given in input, performance are likely to get worse. Based on the Savitzky-Golay filter, we have developed, in this paper, a parallel implementation with engaging a multi-core architecture. Experiments and tests highlight the gain of performance in time terms.

关键词： Performance evaluation filtering Shape Computer architecture filtering algorithms Software Digital filters

来源：评论

学校读者我要写书评

暂无评论

Stacked iterated posterior linearization filter

Stacked iterated posterior linearization filter

引用

International Conference on Information Fusion

作者： Matti Raitoharju Ángel F. García-Fernández Simo Ali-Löytty Simo Särkkä Patria Aviation Tampere University Aalto University Tampere Finland University of Liverpool Liverpool UK Tampere University Tampere Finland Aalto University Helsinki Finland

ISBN: (数字)9781737749769

ISBN: (纸本)9798350371420

The Kalman Filter (KF) is a classical algorithm that was developed for estimating a state that evolves in time based on noisy measurements by assuming linear state transition and measurements models. There exist various KF extensions for non-linear situations, but they are not exact and provide different linearization errors. The Iterated Posterior Linearization Filter (IPLF) does the linearizations iteratively to achieve better linearizations. However, it is possible that some measurements cannot be well linearized using the current knowledge, but their linearization may be better after more measurements are available. Thus, we propose an algorithm that can store the older state elements and measurements when their linearization error is high. The resulting algorithm, the Stacked Iterated Posterior Linearization Filter (S-IPLF), is based on linear dynamic models and uses information from multiple time instances to make the linearization of the measurement function. Results show that the proposed algorithm outperforms traditional KF extensions when some of the measurements cannot be well linearized with the current knowledge, but can be when future information is available.

关键词： Accuracy Smoothing methods Current measurement Heuristic algorithms Estimation Nonlinear filters filtering algorithms Information filters Time measurement Kalman filters

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：