检索结果-内蒙古大学图书馆

TAM: A Computing in Memory based on Tandem Array within STT-MRAM for Energy-Efficient Analog MAC Operation

学校读者我要写书评

暂无评论

TAM: A Computing in Memory based on Tandem Array within STT-...

Design, Automation and Test in Europe Conference and Exhibition

作者： Jinkai Wang Zhengkun Gu Hongyu Wang Zuolei Hao Bojun Zhang Weisheng Zhao Yue Zhang MIIT Key Laboratory of Spintronics Fert Beijing Institute School of Integrated Circuit Science and Engineering Beihang University Beijing China State Key Laboratory of Software Development Environment School of Computer Science and Engineering Beihang University Beijing China Nanoelectronics Science and Technology Center Hefei Innovation Research Institute Beihang University Hefei China

Computing in memory (CIM) has been demonstrated promising for energy efficient computing. However, the dramatic growth of the data scale in neural network processors has aroused a demand for CIM architecture of higher bit density, for which the spin transfer torque magnetic RAM (STT-MRAM) with high bit density and performance arises as an up-and-coming candidate solution. In this work, we propose an analog CIM scheme based on tandem array within STT-MRAM (TAM) to further improve energy efficiency while achieving high bit density. First, the resistance summation based analog MAC operation minimizes the effect of low tunnel magnetoresistance (TMR) by the serial magnetic tunnel junctions (MTJs) structure in the proposed tandem array with smaller area overhead. Moreover, a read scheme of resistive-to-binary is designed to achieve the MAC results accurately and reliably. Besides, the data-dependent error caused by MTJs in series has been eliminated with a proposed dynamic selection circuit. Simulation results of a 2Kb TAM architecture show 113.2 TOPS/W and 63.7 TOPS/W for 4-bit and 8-bit input/weight precision, respectively, and reduction by 39.3% for bit-cell area compared with existing array of MTJs in series.

关键词：

SWin: A Set-Window Division Mechanism to Alleviate I/O Jitters of NVMe SSD Drives

学校读者我要写书评

暂无评论

SWin: A Set-Window Division Mechanism to Alleviate I/O Jitte...

IEEE International Conference on High Performance Computing and Communications (HPCC)

作者： Yuting Liu Liang Wang Jinbin Zhu Runnan Shen Shixuan Jiang Limin Xiao State Key Laboratory of Software Development Environment Beihang University Beijing China School of Computer Science and Engineering Beihang University Beijing China

NVMe SSDs increase the performance of I/O system significantly, but they still suffer from the problem of performance jitters. Although existing studies adopt optimized GC or schedulers to make up for the vulnerabilities, they cannot provide a near “GC-free” performance under various scenarios. In this paper, we propose SWin, a Set-Window division mechanism to alleviate I/O jitters of NVMe SSDs, in which we build an optimized scheduler with cache management on sets. First, we first divide NVMe SSD into set level and determine time windows based on sets. Then, an efficient I/O scheduler is designed to process requests of applications. Finally, we develop a host cache management strategy to ensure a stable response time for I/O requests. The experimental results indicate that the OCSSD system integrated with SWin can gain performance improvement by 32%, 34% and 14% in completion time, IOPS and latency performance, respectively.

关键词： Smart cities Scheduling algorithms Jitter Time factors

ERetinex: Event Camera Meets Retinex Theory for Low-Light Image Enhancement

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Guo, Xuejian Tian, Zhiqiang Wang, Yuehang Li, Siqi Jiang, Yu Du, Shaoyi Gao, Yue the School of Software Engineering Xi’an Jiaotong University Xi’an710049 China the College of Computer Science and Technology Jilin University Changchun130012 China the National Key Laboratory of Human-Machine Hybrid Augmented Intelligence National Engineering Research Center for Visual Information and Applications Institute of Artificial Intelligence and Robotics Xi’an Jiaotong University Xi’an710049 China BNRist THUIBCS KLISS BLBCI School of Software Tsinghua University Beijing100084 China

Low-light image enhancement aims to restore the under-exposure image captured in dark scenarios. Under such scenarios, traditional frame-based cameras may fail to capture the structure and color information due to the exposure time limitation. Event cameras are bio-inspired vision sensors that respond to pixel-wise brightness changes asynchronously. Event cameras’ high dynamic range is pivotal for visual perception in extreme low-light scenarios, surpassing traditional cameras and enabling applications in challenging dark environments. In this paper, inspired by the success of the retinex theory for traditional frame-based low-light image restoration, we introduce the first methods that combine the retinex theory with event cameras and propose a novel retinex-based low-light image restoration framework named ERetinex. Among our contributions, the first is developing a new approach that leverages the high temporal resolution data from event cameras with traditional image information to estimate scene illumination accurately. This method outperforms traditional image-only techniques, especially in low-light environments, by providing more precise lighting information. Additionally, we propose an effective fusion strategy that combines the high dynamic range data from event cameras with the color information of traditional images to enhance image quality. Through this fusion, we can generate clearer and more detail-rich images, maintaining the integrity of visual information even under extreme lighting conditions. The experimental results indicate that our proposed method outperforms state-of-the-art (SOTA) methods, achieving a gain of 1.0613 dB in PSNR while reducing FLOPS by 84.28%. The code is available at https://***/lodew920/ERetinex. © 2025, CC BY.

关键词： Cameras

Fuzzy Neural Network Sliding-Mode Control with Actor-Critic for a Class of Robot Systems

学校读者我要写书评

暂无评论

Fuzzy Neural Network Sliding-Mode Control with Actor-Critic ...

IEEE International Conference on Orange Technologies (ICOT)

作者： Lon-Chen Hung Chun-Cheng Wei Hung-Jen Chen Chih-Fu Yang Wei-Min Cheng Department of Electronic Engineering Lunghwa University of Science and Technology School of Software and Big Data Changzhou College of Information Technology Department of Computer Information and Network Engineering Lunghwa University of Science and Technology

ISBN: (纸本)9781665454209

In this article, a robust fuzzy RBF neural network sliding-mode control with actor-critic for a class of robot systems. Trajectory tracking control of robotic systems has favorable performance for tracking control. The fuzzy RBF neural network sliding-mode and actor-critic method is handled to compensate the uncertainty and disturbance of system. The stability analysis is based on for the proposed adaptive and robust control method. The simulation results show the effectiveness under the uncertainties.

关键词： Uncertainty Target tracking Trajectory tracking Simulation Neural networks Manipulators Stability analysis

An Application-Oriented Cache Allocation and Prefetching Method for Long-Running Applications in Distributed Storage Systems

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2019年第4期28卷 773-780页

作者： GUO Chang LI Ying LIU Hongzhi WU Zhonghai School of Electronics Engineering and Computer Science Peking University National Engineering Research Center For Software Engineering Peking University School of Software and Microelectronics Peking University

Characteristics of long-running applications in cloud and big data environment are various and significantly influence the performance of cache *** gap between existing cache systems and the increasing performance requirements motivates us to propose the Application-oriented cache allocation and prefetching method(ACAP) to improve data access performance. An application-oriented cache allocation approach is designed based on hit count growth rates for a higher overall hit rate. Two application-oriented sequential prefetching approaches are proposed to improve the hit rate and prefetching accuracy by learning average read sizes of long-running applications. Based on correlation of data accesses, a parallelized correlated-directed prefetching approach is proposed to further increase the hit *** approaches are intergrated to obtain the maximized hit rate and prefetching accuracy. Experimental results on12 public real system traces show that ACAP achieves14.03%(up to 33.82%) higher prefetching accuracy and2.01%(up to 7.54%) higher hit rate compared with the best combination of baselines.

关键词： Cache prefetching Cache allocation Distributed storage system Resource management

QUASI-PERRON-FROBENIUS PROPERTY OF A CLASS OF SADDLE POINT MATRICES

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Li, Zheng Zhang, Tie Li, Chang-Jun Department of Mathematics Northeastern University Shenyang110004 China School of Computer and Software Engineering University of Science and Technology Liaoning Anshan114051 China

The saddle point matrices arising from many scientific computing fields have block structure W = (Equation presented), where the sub-block A is symmetric and positive definite, and C is symmetric and semi-nonnegative definite. In this article we report a unobtrusive but potentially theoretically valuable conclusion that under some conditions, especially when C is a zero matrix, the spectral radius of W must be the maximum eigenvalue of W. This characterization approximates to the famous Perron-Frobenius property, and is called quasi-Perron-Frobenius property in this paper. In numerical tests we observe the saddle point matrices derived from some mixed finite element methods for computing the stationary Stokes equation. The numerical results confirm the theoretical analysis, and also indicate that the assumed condition to make the saddle point matrices possess quasi-Perron-Frobenius property is only sufficient rather than *** Codes 65F15, 65F10 © 2022, CC BY-NC-SA.

关键词： Numerical methods

A Scale-Arbitrary Image Super-Resolution Network Using Frequency-domain Information

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Fang, Jing Yu, Yinbo Wang, Zhongyuan Ding, Xin Hu, Ruimin The National Engineering Research Center for Multimedia Software School of Computer Science Wuhan University Wuhan430072 China The School of Cybersecurity Northwestern Polytechnical University Xi’an710072 China The School of Computer and Data Science NingboTech University Ningbo315199 China

Image super-resolution (SR) is a technique to recover lost high-frequency information in low-resolution (LR) images. Spatial-domain information has been widely exploited to implement image SR, so a new trend is to involve frequency-domain information in SR tasks. Besides, image SR is typically application-oriented and various computer vision tasks call for image arbitrary magnification. Therefore, in this paper, we study image features in the frequency domain to design a novel scale-arbitrary image SR network. First, we statistically analyze LR-HR image pairs of several datasets under different scale factors and find that the high-frequency spectra of different images under different scale factors suffer from different degrees of degradation, but the valid low-frequency spectra tend to be retained within a certain distribution range. Then, based on this finding, we devise an adaptive scale-aware feature division mechanism using deep reinforcement learning, which can accurately and adaptively divide the frequency spectrum into the low-frequency part to be retained and the high-frequency one to be recovered. Finally, we design a scale-aware feature recovery module to capture and fuse multi-level features for reconstructing the high-frequency spectrum at arbitrary scale factors. Extensive experiments on public datasets show the superiority of our method compared with state-of-the-art methods. Copyright © 2022, The Authors. All rights reserved.

关键词： Reinforcement learning

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Liu, Yuqing Jia, Qi Zhang, Jian Fan, Xin Wang, Shanshe Ma, Siwei Gao, Wen School of Software Dalian University of Technology Dalian116620 China International School of Information Science and Engineering Dalian University of Technology Dalian116620 China School of Electronic and Computer Engineering Peking University Shenzhen Graduate School Shenzhen China School of Electronics Engineering and Computer Science Institute of Digital Media Peking University Beijing100871 China

As a highly ill-posed issue, single image super-resolution (SISR) has been widely investigated in recent years. The main task of SISR is to recover the information loss caused by the degradation procedure. According to the Nyquist sampling theory, the degradation leads to aliasing effect and makes it hard to restore the correct textures from low-resolution (LR) images. In practice, there are correlations and self-similarities among the adjacent patches in the natural images. This paper considers the self-similarity and proposes a hierarchical image super-resolution network (HSRNet) to suppress the influence of aliasing. We consider the SISR issue in the optimization perspective, and propose an iterative solution pattern based on the half-quadratic splitting (HQS) method. To explore the texture with local image prior, we design a hierarchical exploration block (HEB) and progressive increase the receptive field. Furthermore, multi-level spatial attention (MSA) is devised to obtain the relations of adjacent feature and enhance the high-frequency information, which acts as a crucial role for visual experience. Experimental result shows HSRNet achieves better quantitative and visual performance than other works, and remits the aliasing more effectively. Copyright © 2022, The Authors. All rights reserved.

关键词： Textures

Cross-modal and Semantics-Augmented Asymmetric CycleGAN for Data-Imbalanced Anime Style Face Translation 2021

学校读者我要写书评

暂无评论

Cross-modal and Semantics-Augmented Asymmetric CycleGAN for ...

2021 3rd International Conference on Video, Signal and Image Processing

作者： Shiping Deng Kaoru Uchida Zhengwei Yin School of Software Engineering University of Science and Technology of China China and Graduate School of Computer and Information Sciences Hosei University Japan Graduate School of Computer and Information Sciences Hosei University Japan Alibaba Group China

ISBN: (纸本)9781450385886

Human face to anime face translation has attracted the attention of many researchers in recent years, and various works have achieved high-quality style transfer on conventional tasks. However, existing works often have fatal shortcomings when the target domain training data is heavily insufficient, which is named as imbalanced setting. Here the imbalanced (low-resource) task, generally means there is no sufficient data on the training dataset compared with the conventional task, e.g. the training data size is less than 100. To solve this problem, we propose a multi-modal translation model for a specific style. Based on the cyclic adversarial network and class activation map, we import semantic modality to enhance data information and attention modules, which help the model focus more on the discriminative areas between source and target domain. The experimental results show that our method has superiority in low-resource settings compared with similar existing work.

关键词： image translation low-resource Deep neural networks