检索结果-内蒙古大学图书馆

Data compression Conference (DCC)

作者： Ren, Qingqing Zhu, Shuyong Meng, Xuying Zhang, Yujun Chinese Acad Sci Inst Comp Technol Beijing Peoples R China Univ Chinese Acad Sci Beijing Peoples R China Purple Mt Labs Purple Peoples R China Univ Chinese Acad Sci Nanjing Coll Beijing Peoples R China

ISBN: (纸本)9781665478939

Gradient compression technology has attracted much attention in recent years, due to its high effectiveness in alleviating the communication bottleneck of distributed deep learning. However, except for the communication reduction, it also brings in a significant increase of computational overhead, which limits or even eliminates the communication-reduction benefit brought by gradient compression. To solve the high computational overhead problem, we propose an FPGA-based accelerator for gradient compression in this paper. A high-performance and programmable accelerator architecture is developed for accelerating various gradient compression algorithms by offloading compute-intensive compression operations to FPGA. Also, we design and implement the FPGA-based accelerator based on the popular gradient compression algorithm top-k sparsification. Experimental results show that the new accelerator achieves up to hundreds of times faster than the compression algorithm implemented on CPU and GPU. What's more, the stable and controllable performance under different datasets demonstrates that the proposed accelerator is insensitive to data distribution, which is essential for time sensitive applications.

关键词： Deep learning Accelerator architectures Graphics processing units Data compression compression algorithms Field programmable gate arrays

来源：评论

学校读者我要写书评

暂无评论

Sparsity Order Estimation for Compressed Sensing System Using Sparse Binary Sensing Matrix

引用

IEEE ACCESS 2022年 10卷 33370-33392页

作者： Thiruppathirajan, S. Narayanan, Lakshmi R. Sreelal, S. Manoj, B. S. Indian Space Res Org ISRO Vikram Sarabhai Space Ctr VSSC Thiruvananthapuram 695022 Kerala India Indian Inst Space Sci & Technol IIST Thiruvananthapuram 695547 Kerala India

We present a composite Compressed Sensing system for the acquisition and recovery of compressible signals, where a sparse Binary Sensing Matrix aids Sparsity Order Estimation, and a Gaussian Sensing Matrix aids reconstruction. The Binary Sensing Matrix is deterministic and is adapted according to the varying nature of the sparsity order. We estimate the sparsity order by exploiting the sparse structure of the Binary Sensing Matrix and the statistics of the obtained measurements. We refine the estimates of the sparsity order using a Kalman filter with a discrete Markov model that characterizes the sparsity order variation. A Binary Sensing Matrix-Aided Orthogonal Matching Pursuit is developed for faster recovery of compressible signals. Simulation results on real-world and synthetic data demonstrate the merits of the proposed sparsity order estimation and recovery methods compared to other existing methods. Our proposed methods are practical and recover compressible signals at least 25% faster than the existing methods.

关键词： Sensors Sparse matrices Matching pursuit algorithms Estimation Size measurement Noise measurement Volume measurement Compressed sensing compression algorithms discrete cosine transforms greedy algorithms Kalman filters Markov processes maximum likelihood estimation signal reconstruction sparse matrices vibration measurement

来源：评论

学校读者我要写书评

暂无评论

Lossy compression of Quality Values in Sequencing Data

引用

IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS 2021年第5期18卷 1958-1969页

作者： Morales, Veronica Suaste Houghten, Sheridan CONABIO Mexico City 14010 DF Mexico Brock Univ Dept Comp Sci St Catharines ON L2S 3A1 Canada

The dropping cost of sequencing human DNA has allowed for fast development of several projects around the world generating huge amounts of DNA sequencing data. This deluge of data has run up against limited storage space, a problem that researchers are trying to solve through compression techniques. In this study we address the compression of SAM files, the standard output files for DNA alignment. We specifically study lossy compression techniques used for quality values reported in the SAM file and analyze the impact of such lossy techniques on the CRAM format. We present a series of experiments using a data set corresponding to individual NA12878 with three different fold coverages. We introduce a new lossy model, dynamic binning, and compare its performance to other lossy techniques, namely Illumina binning, LEON and QVZ. We analyze the compression ratio when using CRAM and also study the impact of the lossy techniques on SNP calling. Our results show that lossy techniques allow a better CRAM compression ratio. Furthermore, we show that SNP calling performance is not negatively affected and may even be boosted.

关键词： Sequential analysis DNA Standards Genomics Bioinformatics Data models compression algorithms compression next-generation sequencing quality values SNP calling performance

来源：评论

学校读者我要写书评

暂无评论

Modified Adaptive Pulse compression Algorithm for Targets With Range-Straddling

引用

IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS 2021年第5期57卷 3057-3070页

作者： Ning, Chen Tian, Jing Li, Kun Wu, Siliang Beijing Inst Technol Sch Informat & Elect Beijing 100081 Peoples R China

Adaptive pulse compression (APC) has been proposed based on the minimum mean square error (MMSE) criterion to effectively suppress range sidelobes of strong targets and retrieve all targets in multi-target scenarios. However, MMSE-based algorithms suffer from deteriorated performance in the presence of targets with range-straddling due to modelling mismatch. This degradation can be partially compensated by modified MMSE-based algorithms. To further suppress sidelobes when range straddling occurs, we propose a modified APC algorithm robust to targets with range-straddling, namely range-straddling-robust APC (RSR-APC). We first establish the signal model for targets with range-straddling and then derive the expressions of the MMSE filter based on matched filter outputs and the range-straddling offsets based on an improved Rife algorithm. Simulation results show that the proposed algorithm can suppress sidelobes of targets with range-straddling and improve estimation accuracy of target positions in different scenarios.

关键词： Estimation Adaptation models compression algorithms Simulation Signal to noise ratio Mean square error methods Computational modeling Adaptive pulse compression (APC) minimum mean square error range straddling sidelobe suppression

来源：评论

学校读者我要写书评

暂无评论

Adaptive Data compression for Classification Problems

引用

IEEE ACCESS 2021年 9卷 157654-157669页

作者： Pourkamali-Anaraki, Farhad Bennette, Walter D. Univ Massachusetts Dept Comp Sci Lowell MA 01854 USA Air Force Res Lab Informat Directorate Rome NY 13441 USA

Data subset selection is a crucial task in deploying machine learning algorithms under strict constraints regarding memory and computation resources. Despite extensive research in this area, a practical difficulty is the lack of rigorous strategies for identifying the optimal size of the reduced data to regulate trade-offs between accuracy and efficiency. Furthermore, existing methods are often built around specific machine learning models, and translating existing theoretical results into practice is challenging for practitioners. To address these problems, we propose two adaptive compression algorithms for classification problems by formulating data subset selection in the form of interactive teaching. The user interacts with the learning task at hand to adapt to the unique structure of the problem at hand, developing an iterative importance sampling scheme. We also propose to couple importance sampling and a diversity criterion to further control the evolution of the data summary over the rounds of interaction. We conduct extensive experiments on several data sets, including imbalanced and multiclass data, and various classification algorithms, such as ensemble learning and neural networks. Our results demonstrate the performance, efficiency, and ease of implementation of the underlying framework.

关键词： Data compression Data models Training data Neural networks Computational modeling Training Task analysis Adaptive algorithms supervised learning classification algorithms compression algorithms

来源：评论

学校读者我要写书评

暂无评论

Fast Neural Representations for Direct Volume Rendering

引用

COMPUTER GRAPHICS FORUM 2022年第6期41卷 196-211页

作者： Weiss, S. Hermueller, P. Westermann, R. Tech Univ Munich Munich Germany

Despite the potential of neural scene representations to effectively compress 3D scalar fields at high reconstruction quality, the computational complexity of the training and data reconstruction step using scene representation networks limits their use in practical applications. In this paper, we analyse whether scene representation networks can be modified to reduce these limitations and whether such architectures can also be used for temporal reconstruction tasks. We propose a novel design of scene representation networks using GPU tensor cores to integrate the reconstruction seamlessly into on-chip raytracing kernels, and compare the quality and performance of this network to alternative network- and non-network-based compression schemes. The results indicate competitive quality of our design at high compression rates, and significantly faster decoding times and lower memory consumption during data reconstruction. We investigate how density gradients can be computed using the network and show an extension where density, gradient and curvature are predicted jointly. As an alternative to spatial super-resolution approaches for time-varying fields, we propose a solution that builds upon latent-space interpolation to enable random access reconstruction at arbitrary granularity. We summarize our findings in the form of an assessment of the strengths and limitations of scene representation networks for compression domain volume rendering, and outline future research directions. Source code:

关键词： volume rendering neural networks compression algorithms

来源：评论

学校读者我要写书评

暂无评论

IBU: An In-Block Update Address Mapping Scheme for Solid-State Drives

引用

IEEE ACCESS 2022年 10卷 4934-4947页

作者： Taghizadeh, Reza Gholami Marvasti, Mohammadreza Binesh Asghari, Seyyed Amir Taghizadeh, Ramin Gholami Nabavi, Morteza Savaria, Yvon Kharazmi Univ Dept Elect & Comp Engn Tehran *** Iran Polytech Montreal Dept Elect & Comp Engn Montreal PQ H3T 1J4 Canada

One of the most important features of solid-state drives (SSDs) is the ability to update data sectors out of place, performing garbage collection operations within its physical blocks, to mitigate limited flash memory lifespan. In SSDs, a firmware called the Flash Translation Layer (FTL) is used to hide such features from the operating system. One way to extend the lifespan of flash memories is to use data compression at the FTL level. In compression-based FTLs, the address mapping type has a significant effect on the efficiency and read/write speed. The type of address mapping in previous similar schemes is based on page-level address translation and requires high memory space for the mapping table. In this paper, we propose a compression-based hybrid FTL called In Block Update Address Mapping, in which the pages of each logical block are compressed, and fewer physical pages are written in physical blocks. Therefore, a small number of free pages remain. In the proposed scheme, the required memory of the mapping table has been reduced by 78% when compared to similar schemes. Moreover, extensive simulation results reported in this paper show that the proposed FTL scheme outperforms other FTL schemes in the read and write operations under realistic benchmark workloads.

关键词： Flash memories Data compression Writing Memory management Solid state drives Hardware compression algorithms Flash memory SSD flash translation layer garbage collection data compression in SSD

来源：评论

学校读者我要写书评

暂无评论

Distillation-Guided Residual Learning for Binary Convolutional Neural Networks

引用

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 2022年第12期33卷 7765-7777页

作者： Ye, Jianming Wang, Jingdong Zhang, Shiliang Peking Univ Dept Comp Sci Sch Elect Engn & Comp Sci Beijing 100871 Peoples R China Microsoft Res Asia Beijing 100080 Peoples R China

It is challenging to bridge the performance gap between binary convolutional neural network (BCNN) and floating-point CNN (FCNN). This performance gap is mainly caused by the inferior modeling capability and training strategy of BCNN, which leads to substantial residuals in intermediate feature maps between BCNN and FCNN. To minimize the performance gap, we enforce BCNN to produce similar intermediate feature maps with the ones of FCNN. This intuition leads to a more effective training strategy for BCNN, i.e., optimizing each binary convolutional block with blockwise distillation loss derived from FCNN. The goal of minimizing the residuals in intermediate feature maps also motivates us to update the binary convolutional block architecture to facilitate the optimization of blockwise distillation loss. Specifically, a lightweight shortcut branch is inserted into each binary convolutional block to complement residuals at each block. Benefited from its squeeze-and-interaction (SI) structure, this shortcut branch introduces a fraction of parameters, e.g., less than 10% overheads, but effectively boosts the modeling capability of binary convolution blocks in BCNN. Extensive experiments on ImageNet demonstrate the superior performance of our method in both classification efficiency and accuracy, e.g., BCNN trained with our methods achieves the accuracy of 60.45% on ImageNet, better than many state-of-the-art ones.

关键词： Training Convolution Optimization Kernel Computer architecture Computational modeling Neural networks Binary neural network compression algorithms distillation

来源：评论

学校读者我要写书评

暂无评论

An Adaptive Low-Complexity compression Algorithm for Vehicle Systems

An Adaptive Low-Complexity Compression Algorithm for Vehicle...

引用

Consumer Technology (ICCT-Pacific), International Conference on

作者： Chun-Hsiang Huang Shih-Lun Chen He-Sheng Chou Ya-Yun Huang Department of Electronic Engineering Chung Yuan Christian University Taoyuan Taiwan Department of Program on Semiconductor Manufacturing Technology Academy of Innovative Semiconductor and Sustainable Manufacturing National Cheng Kung University Tainan Taiwan

ISBN: (数字)9798331504120

ISBN: (纸本)9798331504137

Advanced Driver Assistance Systems (ADAS) are increasingly integrated into modern vehicles, relying on high-speed image transmission from on-board cameras to control systems. This study presents an adaptive lossless image compression algorithm designed to reduce transmission data while maintaining high image quality. By dynamically switching between Color Filter Array (CFA) compression and predictive compression, the proposed method achieves an average compression rate of 3.835 and a Peak Signal-to-Noise Ratio (PSNR) of 30.57 dB, significant improvements in efficiency, making it suitable for automotive applications.

关键词： Image quality Image coding PSNR Image color analysis Heuristic algorithms Automotive applications Filtering algorithms Prediction algorithms Real-time systems compression algorithms

来源：评论

学校读者我要写书评

暂无评论

LZHV: Accelerating LZ77 compression Algorithm with Hash Verification

LZHV: Accelerating LZ77 Compression Algorithm with Hash Veri...

引用

Data compression Conference (DCC)

作者： Guodong Ye Tianyi Wang Xin Liu Rui Zhou Qingguo Zhou School of Information Science & Engineering Lanzhou University Lanzhou Gansu China

ISBN: (数字)9798331534714

ISBN: (纸本)9798331534721

The longest match strategy in LZ77, a major bottleneck in the compression process, is accelerated in enhanced algorithms such as LZ4 and ZSTD by using a hash table. However, it may results in numerous random memory accesses, which modern CPUs handle inefficiently, thus reducing the compression speed. In this paper, we introduce the LZHV algorithm, which significantly reduces unnecessary memory accesses by over 99% through the implementation of hash verification within the hash table. By integrating LZHV into LZ4 at its default compression level and ZSTD at levels 3 and 4, we achieve a compression speed improvement of over 10% across various platforms.

关键词： Data compression Optimization compression algorithms

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：