检索结果-内蒙古大学图书馆

4th Annual International Conference on Multimedia Retrieval (ICMR)

作者： Deng, Ziqing Lai, Zhihui Ding, Yujuan Kong, Heng Wu, Xu Shenzhen Univ Coll Comp Sci & Software Engn Shenzhen Peoples R China Hong Kong Polytech Univ Dept Comp Shenzhen Peoples R China BaoAn Cent Hosp Shenzhen Dept Breast & Thyroid Surg Shenzhen Peoples R China

ISBN: (纸本)9798400706028

Hash learning aims to map multimedia data into Hamming space, in which the data point is represented by low-dimensional binary codes and the similarity relationships are preserved. Despite existing hash learning methods have been effectively used in data retrieval tasks for its merits of low memory cost and high computational efficiency, there still remain two major technical challenges. Firstly, due to the discrete constraints of hash codes, traditional hash methods typically use relaxation strategy to learn real-value features and then quantize them into binary codes through a sign function, resulting in significant quantization errors. Secondly, hash codes are usually low-dimensional, which would be inadequate to preserve either the information of each data point or the relationship between two. These two challenges would greatly limit the retrieval performance of learned hash codes. To solve these problems, we introduce a novel quantization method called scaling factor quantization to enhance hash learning. Unlike traditional hashing methods, we propose to map the data into two parts, i.e., hash codes and scaling factors, to learn the representative codes for the use of retrieval. Specifically, we design a multi-output branch network structure, i.e., Deep Scaling factor quantization Network (DSQN) and an iterative training strategy for DSQN to learn the two parts of mapping. Comprehensive experiments conducted on three benchmark datasets demonstrate that the hash codes and scaling factors learned by DSQN significantly improve retrieval accuracy compared to existing hash learning methods.

关键词： Hash learning quantization errors Scaling factor

来源：评论

学校读者我要写书评

暂无评论

High Precision Timing Based on Parabola Curve Fitting in Narrowband System

High Precision Timing Based on Parabola Curve Fitting in Nar...

引用

International Conference on Cyber-Enabled Distributed Computing and Knowledge Discovery (CyberC)

作者： Zou, Jun Xu, Chen Nanjing Univ Sci & Technol Nanjing Peoples R China

ISBN: (纸本)9781728125428

An accurate timing is the prerequisite for establishing communication link. The arbitrarily of the signal's arrival time will lead to sampling quantization errors. In broadband system now widely applied, quantization errors are limited in allowable range thanks to relatively high sampling rate. However, sampling rate in narrowband system is much lower. Increasing sampling rate will add complexity, so a better solution is in urgent need. In this paper, we firstly deduce the mathematical equation of the direct sliding correlator, whose input is a synchronization sequence with sampling delay. After that we analyze the fitting characteristic of the correlator output curve, based on this a feasible parabola curve fitting method is proposed. Finally, the simulation results demonstrate that our proposed optimization method can work better than direct sliding correlator even at a lower sampling rate.

关键词： quantization errors timing detection narrowband curve fitting

来源：评论

学校读者我要写书评

暂无评论

Accuracy considerations in the design of digital instrumentation

引用

Measurement 1991年第3期9卷 124-133页

作者： Kästli, U Swiss Federal Institute of Technology Zurich Switzerland

The most important effects of operations specific to digital signal processing - sampling, quantising and windowing of a data set - and their influence on accuracy of digital instrumentation for electrical power systems, are presented. As an evaluation base, a simple power line signal model is introduced. A reference A/D converter system, suitable for experimental verification purposes, is described. Oversampling, dithering and autocalibration techniques do guarantee elevated accuracy. A high-resolution sampling-frequency-generator permits synchronous operation to the measurement signal frequency. © 1991.

关键词： digital measurement dither oversampling Power line signal model quantisation errors quantization errors

来源：评论

学校读者我要写书评

暂无评论

A fixed-point implementation of explicit MPC laws

A fixed-point implementation of explicit MPC laws

引用

American Control Conference

作者： Darup, Moritz Schulze Redder, Adrian Quevedo, Daniel Univ Paderborn Dept Elect Engn Automat Control Grp Paderborn Germany

ISBN: (纸本)9781538654286

We present a method for the implementation of predictive controllers for linear constrained systems based on fixed-point numbers. Fixed-point implementations are of high practical interest since they are fast and since cheap processing units with low energy consumption are available. Our scheme builds on the assumption that an explicit description of the predictive control law, which is known to be piecewise affine, is at hand. In principle, an implementation of the controller is then straightforward. However, fixed-point representations of the control law and the state measurements cause quantization errors that may affect the stability of the system. We show how robust MPC can be used to compensate those quantization errors and to guarantee robust stability.

关键词： Fixed-point numbers quantization errors robust MPC linear constrained systems

来源：评论

学校读者我要写书评

暂无评论

Multi-Wavelength Parallel Training and quantization-Aware Tuning for WDM-Based Optical Convolutional Neural Networks ConsideringWavelength-Relative Deviations 23

Multi-Wavelength Parallel Training and Quantization-Aware Tu...

引用

28th Asia and South Pacific Design Automation Conference (ASP-DAC)

作者： Zhu, Ying Liu, Min Xu, Lu Wang, Lei Xiao, Xi Yu, Shaohua CICT State Key Lab OCTN Wuhan Peoples R China NOEIC CICT Wuhan Peoples R China

ISBN: (纸本)9781450397834

Wavelength Division Multiplexing (WDM)-based Mach-Zehnder Interferometer Optical Convolutional Neural Networks (MZI-OCNNs) have emerged as a promising platform to accelerate convolutions that cost most computing sources in neural networks. However, the wavelength-relative imperfect split ratios and actual phase shifts in MZIs and quantization errors from the electronic configuration module will degrade the inference accuracy of WDM-based MZIOCNNs and thus render them unusable in practice. In this paper, we propose a framework that models the split ratios and phase shifts under different wavelengths, incorporates them into OCNN training, and introduces quantization-aware tuning to maintain inference accuracy and reduce electronic module complexity. Consequently, the framework can improve the inference accuracy by 49%, 76%, and 76%, respectively, for LeNet5, VGG7, and VGG8 implemented with multi-wavelength parallel computing. And instead of using Float 32/ 64 quantization resolutions, only 5,6, and 4 bits are needed and fewer quantization levels are utilized for configuration signals.

关键词： CNN accelerator WDM device deviations quantization errors

来源：评论

学校读者我要写书评

暂无评论

A new error-mapping scheme for scalable audio coding 16

A new error-mapping scheme for scalable audio coding

引用

2014 16th IEEE International Workshop on Multimedia Signal Processing, MMSP 2014

作者： Huang, Haibin Rahardja, Susanto Institute for Infocomm Research 1 Fusionopolis Way Singapore Singapore Department of Electrical and Computer Engineering National University of Singapore 21 Lower Kent Ridge Road Singapore Singapore

ISBN: (纸本)9781479958962

In scalable audio coders, such as the MPEG-4 SLS, error-mapping is used to map quantization errors in the core coder to an error signal before passing through bit-plane coding. In this paper, we propose a new error-mapping scheme that is derived by observing statistical properties of the error signal. Compared with the error-mapping in SLS, the proposed scheme improves coding efficiency as well as computational complexity of the coder. An average improvement of 9 points in MUSHRA score has been achieved by the proposed scheme in subjective listening tests. The proposed error-mapping adds a useful new tool to the existing toolset for constructing next-generation scalable audio coders.

关键词： audio coding codecs computational complexity quantisation (signal) MPEG-4 SLS MUSHRA audio coders bit-plane coding error-mapping scheme quantization errors scalable audio coding Audio coding Conferences Decoding Histograms quantization (signal) Transform coding Audio coding bit-plane coding complexity classes quantization transform coding codecs histograms DECODING ERROR SIGNAL ENCODER

来源：评论

学校读者我要写书评

暂无评论

Numerical quantization effects in KHILS projection systems

Numerical quantization effects in KHILS projection systems

引用

Conference on Technologies for Synthetic Environments - Hardware-in-the-Loop Testing VII

作者： Flynn, DS Thompson, RA Goldsmith, GC MacAulay Brown Inc Shalimar FL 32579 USA

ISBN: (纸本)0819444677

The Honeywell resistor arrays produce radiance outputs, which are observed to have a strong non-linear dependence on the voltage out of the digital-to-analog-converters (DACs). In order for the projection system to run in a radiometrically calibrated mode, the radiances in the image generator must be transformed with exactly the inverse of the resistor array response function before they are sent to the DACs. Representing the image values out of the image generator and the values into the DACs with quantized, digital values introduces errors in the radiance out of the resistor array. Given the functional form of the emitter array response and the number of bits used to represent the image values, these errors in the radiometric output due to the quantization effects can be calculated. This paper describes the calculations and presents results for WISP, MSSP, and the new extended range and standard range BRITE II arrays.

关键词： quantization errors radiometric accuracy radiometric resolution dynamic range KHILS BRITE II WISP

来源：评论

学校读者我要写书评

暂无评论

METROLOGICAL PROPERTIES OF A/D CONVERTERS UTILIZING HIGHER ORDER SIGMA-DELTA MODULATORS COMPARED WITH A/D CONVERTERS WITH MODULATORS OF FIRST ORDER

引用

METROLOGY AND MEASUREMENT SYSTEMS 2014年第1期21卷 37-46页

作者： Sidor, Tadeusz AGH Univ Sci & Technol Fac Elect Engn Automat Comp Sci & Biomed Engn Krakow Poland

Time domain analysis is used to determine whether A/D converters that employ higher order sigma-delta modulators, widely used in digital acoustic systems, have superior performance over classical synchronous A/D converters with modulators of first order when taking into account their important metrological property which is the magnitude of the quantization error. It is shown that the quantization errors of delta-sigma A/D converters with higher order modulators are exactly on the same level as for converters with a first order modulator.

关键词： A/D converters sigma-delta modulators quantization errors

来源：评论

学校读者我要写书评

暂无评论

Automated calibration technique for programmable Universal Frequency-to-Digital Converter

引用

IFAC Proceedings Volumes 2006年第21期39卷 72-77页

作者： Sergey Y. Yurish Nikolay V. Kirianaki National University Lviv Polytechmic Bandera str. 12 Lviv UA International Frequency Sensor Association (IFSA).

The automated calibration technique for a trimming inaccuracy (calibrationtolerance) of quartz resonator using in the universal precision frequency-to-digital converter is described in the paper. It was shown that such technique can be used also for reducing an error due to a quartz crystal aging and a temperature drift. The experimental technique and results for converter's error research in the temperature range -40…+85 °C is also described in the paper. The aim of these experiments was to detennine the margins of temperature range outside of which it is necessary to use the correction of frequency instability due to tcmperature changes.

关键词： calibration oscillators frequency conversion frequency measurements tolerance quantization errors converters

来源：评论

学校读者我要写书评

暂无评论

Optimal Structures for Real-time Implementing Linear Digital State-Space System

引用

IFAC Proceedings Volumes 1995年第5期28卷 427-434页

作者： S. Hadri C. Iung Centre de Recherche en Automatique de Nancy 2 avenue Forêt de Haye 54516 Vandoeuvre-Les-Nancy Cedex.FRANCE Fax 83595644

In this paper, a numerical robustness of digital system structure is analysed, in forward manner, by means of a theoretical quantization errors propagation study of both roundoff noise and coefficient quantization errors. This study takes departure from the description of the quantized state-space. It is shown that the previous methods for optimal structures synthesis are not suitable and leads to restraint conditions of optimality. Then more general conditions and optimal realizations are derived.

关键词： digital systems state-space realization quantization errors error propagation optimal structures

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：