检索结果-内蒙古大学图书馆

IEEE International conference on Big Data and Cloud Computing (BdCloud)

作者： Yuting Zhu Hongxu Jiang Runhua Zhang Yonghua Zhang Dong Dong Beijing Key Laboratory of Digital Media of Beihang University Beihang University Beijing China

Channel pruning is one of the mainly used meth-ods in current network model compression. However, existing channel pruning methods lack effective hardware runtime latency guidance, making the reduction in model size not fully converted into inference latency reduction, which results in poor inference performance of the pruned network models. This paper proposes a CNN channel pruning framework, CNNBooster, that incorporates hardware runtime latency in-formation, whose two major contributions are as follows: First, CNNBooster automatically analyses the latency behavior of various CNNs with channel reduction on different GPU hard-ware platforms, which achieves efficient localization of pruna-ble coordinates. Second, CNNBooster uses a flexible grained latency-aware and param-aware pruning algorithm based on prunable coordinates to achieve a significant reduction in model inference latency and parameter size. We evaluate CNNBooster with three benchmark models (VGG16, ResNet18 and MobilenetV1) on three NVIDIA-series platforms (V100, RTX 2080 Ti and Jetson Nano). The experimental results show that CNNBooster can achieve a maximum 67.51 % latency re-duction and 1.4x performance improvement compare with current state-of-the-art works.

关键词： Location awareness Runtime Graphics processing units Benchmark testing Hardware Inference algorithms Behavioral sciences

来源：评论

学校读者我要写书评

暂无评论

Maximum Mean and Covariance Discrepancy for Unsupervised Domain Adaptation

引用

NEURAL processing LETTERS 2020年第1期51卷 347-366页

作者： Zhang, Wenju Zhang, Xiang Lan, Long Luo, Zhigang Natl Univ Def Technol Sci & Technol Parallel & Distributed Lab Changsha 410073 Peoples R China Natl Univ Def Technol State Key Lab High Performance Comp Changsha 410073 Peoples R China Natl Univ Def Technol Coll Comp Changsha 410073 Peoples R China

A fundamental research topic in domain adaptation is how best to evaluate the distribution discrepancy across domains. The maximum mean discrepancy (MMD) is one of the most commonly used statistical distances in this field. However, information about distributions could be lost when adopting non-characteristic kernels by MMD. To address this issue, we devise a new distribution metric named maximum mean and covariance discrepancy (MMCD) by combining MMD and the proposed maximum covariance discrepancy (MCD). MCD probes the second-order statistics in reproducing kernel Hilbert space, which equips MMCD to capture more information compared to MMD alone. To verify the efficacy of MMCD, an unsupervised learning model based on MMCD abbreviated as McDA was proposed and efficiently optimized to resolve the domain adaptation problem. Experiments on image classification conducted on two benchmark datasets show that McDA outperforms other representative domain adaptation methods, which implies the effectiveness of MMCD in domain adaptation.

关键词： Domain adaptation image classification Dimensionality reduction Transfer learning

来源：评论

学校读者我要写书评

暂无评论

Network Dynamics Based Sensor Data processing 30

Network Dynamics Based Sensor Data Processing

引用

conference on Automatic Target Recognition XXX

作者： Li, Bingcheng Lockheed Martin 1801 State Route 17CMD0315 Owego NY 13827 USA

ISBN: (数字)9781510635661

ISBN: (纸本)9781510635661;9781510635654

Two-dimensional (2D) image processing and three-dimensional (3D) LIDAR point cloud data analytics are two important techniques of sensor data processing for many applications such as autonomous systems, auto driving cars, medical imaging and many other fields. However, 2D image data are the data that are distributed in regular 2D grids while 3D LIDAR data are represented in point cloud format that consist of points nonuniformly distributed in 3D spaces. Their different data representations lead to different data processing techniques. Usually, the irregular structures of 3D LIDAR data often cause challenges of 3D LIDAR analytics. Thus, very successful diffusion equation methods for image processing are not able to apply to 3D LIDAR processing. In this paper, applying network and network dynamics theory to 2D images and 3D LIDAR analytics, we propose graph-based data processing techniques that unify 2D image processing and 3D LIDAR data analytics. We demonstrate that both 2D images and 3D point cloud data can be processed in the same framework, and the only difference is the way to choose neighbor nodes. Thus, the diffusion equation techniques in 2D image processing can be used to process 3D point cloud data. With this general framework, we propose a new adaptive diffusion equation technique for data processing and show with experiments that this new technique can perform data processing with high performance.

关键词： image processing Diffusion Anisotropic diffusion 3D image processing Data modeling Data processing Sensors LIDAR Clouds Analytics

来源：评论

学校读者我要写书评

暂无评论

Multiple Transform Selection concept modeling and implementation using Interface Based SDF graphs 21

Multiple Transform Selection concept modeling and implementa...

引用

14th Workshop on Design and Architectures for Signal and image processing, DASIP 2021, held jointly with the 16th HiPEAC conference

作者： Haggui, Naouel Belghith, Fatma Hamidouche, Wassim Masmoudi, Nouri Nezan, Jean-François Ietr Laboratory Insa Rennes France National School of Engineering of Sfax Tunisia

ISBN: (纸本)9781450389013

Recent studies predict that video data accounts for 82% of Internet traffic by 2022. This fact has motivated MPEG to define a new Video Coding Standard called Versatile Video Coding (VVC), which will be released by the end of 2020. VVC will offer the possibility to handle new video formats and to improve significantly video compression over its predecessor HEVC. Indeed, the objective is to reduce the necessary bit rate by half, at equivalent quality. These advances require the use of more complex algorithms, although the increase in complexity has been limited throughout the standardization process. In order to decrease the complexity of VVC and consequently the coding execution time, several methods have been introduced at different stages of the encoder. The aim of this paper is to explore the available parallelism of VVC to accelerate the coding and the decoding processes. This paper focuses on the transformation block and more specifically the new concept of Multiple Transform Selection (MTS) introduced by VVC. Moreover, a study of several granularity levels of Interface-Based Synchronous Dataflow (IBSDF) models and their impact on the performances obtained on x86 architectures is presented. IBSDF dataflow graph has been developed to reveal the available parallelism of MTS. The PREESM fast prototyping tool is then used for the mapping and the scheduling of MTS on virtual and real parallel architectures and for generating efficient parallel implementations on real architectures. PREESM has been used in this work to explore the potential parallelism offered by MTS and to prove the efficiency of MTS on multicore x86 architectures. Experimental results show a speed-up close to the optimum. © 2021 ACM.

关键词： Video signal processing

来源：评论

学校读者我要写书评

暂无评论

An Embedded Vision System for Autonomous Vehicles

An Embedded Vision System for Autonomous Vehicles

引用

conference on Industrial Optical Devices and Systems (ODS)

作者： Shehaz, Faisal Univ Bremen Elect & Elect Engn Dept Bremen Germany

ISBN: (纸本)9781510638075

In this paper, a real-time perception system for autonomous car is presented. It is based on a highly parallel architecture using state of the art Field Programmable Gate Array (FPGA) to perform both low and intermediary levels image processing tasks at video frame rate ( i.e. 30 frames / s). The hardware algorithm consists to perform noise removal and edge detection, followed by Hough transform task to extract the segments corresponding the lanes boundaries. The rich hardware resources which are available in nowadays FPGAs (e.g. large built-in distributed RAM memories, DSP blocks, and reconfigurable PLLs) yielded for a compact and low power consumption real-time vision system. Series of tests on different roads within Abu Dhabi city were successfully conducted for different scenarios such as continues lines, discontinues lines and slightly curved lines for which the car speed reached up to 122 km/h.

关键词： image processing Roads Field programmable gate arrays Video Video processing Computer architecture Digital signal processing Edge detection Unmanned vehicles

来源：评论

学校读者我要写书评

暂无评论

Differential games and methods of digital image processing 6

Differential games and methods of digital image processing

引用

6th Annual International conference on Network and Information Systems for Computers, ICNISC 2020

作者： Mamatov, Mashrabjon Alimov, Xakimjon Department of Geometry and Topology National University of Uzbekistan Named After M. Ulugbek University town Tashkent100174 Uzbekistan Department of Mathematics Jizzakh State Pedagogical Institute Jizzakh130100 Uzbekistan

In this paper, we study the method of digital image processing as a game task with distributed parameters. The possibilities of formulating digital processing as a differential game of the Dirichlet problem for the Poisson equation are indicated. Replacing the differential game with a discrete game with boundary conditions, sufficient conditions are obtained for the possibility of completing the pursuit. In solving the pursuit problem, Chebyshev polynomials of the second kind are used. © 2020 Journal of Physics: conference Series.

关键词： Poisson equation

来源：评论

学校读者我要写书评

暂无评论

Design of Hardware Acceleration of Matched Filter for Retinal Vessel Enhancement

Design of Hardware Acceleration of Matched Filter for Retina...

引用

IEEE International conference on Automation, Electronics and Electrical Engineering (AUTEEE)

作者： Yueying Zhu Wenke Ji Chenfei Liao Nanjing University of Aeronautics and Astronautics Nanjing China Xidian University Xi'an China China University of Mining & Technology Beijing Beijing China

ISBN: (数字)9781665471978

ISBN: (纸本)9781665471985

Clinically, retinal image characterization is utilized for the early diagnosis and treatment of retinal illness. It is necessary to automatically segregate vascular pictures from the retina and perform real-time image augmentation. image enhancement methods based on matching filters and FPGAs were previously available. This study is an expansion of past work. The hardware acceleration of matched filtering extends from a one-dimensional timing signal to a two-dimensional picture space. This article configures 17*17 matched filter kernels in 12 directions, applies pixel stream processing mode with shift registers for image data reading, calculates the convolution of the original image data with each filter kernel, and then takes the largest term as the response output for storage. This saves FPGA memory and accelerates the convolution module based on the parallel operation. In addition, compared to the store-and-read strategy, the design lowers resource and time loss in the data readout module. This article demonstrates how to configure the matched filter cores, read the input, compute, output, and save the results.

关键词： Matched filters image segmentation image recognition Convolution Blood vessels Retina Shift registers

来源：评论

学校读者我要写书评

暂无评论

Proceedings of the International conference on New Trends in Signal processing, NTSP 2020

Proceedings of the International Conference on New Trends in...

引用

2020 International conference on New Trends in Signal processing, NTSP 2020

ISBN: (纸本)9781728161556

The proceedings contain 28 papers. The topics discussed include: x-controlled memristive devices for automatic gain control in RC oscillators;analysis of influence of atmospheric conditions on classification DVB-S2 signals;active RC high order filters suitable for anti-aliasing and/or reconstruction filters;processing the RGB images to generate mezzotint plate and simulate the press;entropy based image quality assessment of stego images created by pulse coupled neural network;parallel image signal processing in a distributed car plate recognition system;development of speaker recognizer using I-vectors in two programming environments;finding the sensitivity to transfer branch by graphs;prediction of wind turbines Doppler frequency shifts;and modeling output signals of solid-state photomultiplier with capacitive coupling.

关键词：

来源：评论

学校读者我要写书评

暂无评论

DiffInfinite: large mask-image synthesis via parallel random patch diffusion in histopathology 23

DiffInfinite: large mask-image synthesis via parallel random...

引用

Proceedings of the 37th International conference on Neural Information processing Systems

作者： Marco Aversa Gabriel Nobis Miriam Hägele Kai Standvoss Mihaela Chirica Roderick Murray-Smith Ahmed Alaa Lukas Ruff Daniela Ivanova Wojciech Samek Frederick Klauschen Bruno Sanguinetti Luis Oala University of Glasgow and Dotphoton Glasgow United Kingdom Fraunhofer HHI Berlin Germany Aignostics Berlin Germany Institute of Pathology LMU Munich Munich Germany University of Glasgow Glasgow United Kingdom UC Berkeley Berkeley California Fraunhofer HHI and TU Berlin Berlin Germany Dotphoton Zug Switzerland

We present DiffInfinite, a hierarchical diffusion model that generates arbitrarily large histological images while preserving long-range correlation structural information. Our approach first generates synthetic segmentation masks, subsequently used as conditions for the high-fidelity generative diffusion process. The proposed sampling method can be scaled up to any desired image size while only requiring small patches for fast training. Moreover, it can be parallelized more efficiently than previous large-content generation methods while avoiding tiling artifacts. The training leverages classifier-free guidance to augment a small, sparsely annotated dataset with unlabelled data. Our method alleviates unique challenges in histopathological imaging practice: large-scale information, costly manual annotation, and protective data handling. The biological plausibility of DiffInfinite data is evaluated in a survey by ten experienced pathologists as well as a downstream classification and segmentation task. Samples from the model score strongly on anti-copying metrics which is relevant for the protection of patient data.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Learning with convolution and pooling operations in kernel methods 22

Learning with convolution and pooling operations in kernel m...

引用

Proceedings of the 36th International conference on Neural Information processing Systems

作者： Theodor Misiakiewicz Song Mei Department of Statistics Stanford University Stanford CA Department of Statistics University of California Berkeley Berkeley CA

ISBN: (纸本)9781713871088

Recent empirical work has shown that hierarchical convolutional kernels inspired by convolutional neural networks (CNNs) significantly improve the performance of kernel methods in image classification tasks. A widely accepted explanation for their success is that these architectures encode hypothesis classes that are suitable for natural images. However, understanding the precise interplay between approximation and generalization in convolutional architectures remains a challenge. In this paper, we consider the stylized setting of covariates (image pixels) uniformly distributed on the hypercube, and characterize exactly the RKHS of kernels composed of single layers of convolution, pooling, and downsampling operations. We use this characterization to compute sharp asymptotics of the generalization error for any given function in high-dimension. In particular, we quantify the gain in sample complexity brought by enforcing locality with the convolution operation and approximate translation invariance with average pooling. Notably, these results provide a precise description of how convolution and pooling operations trade off approximation with generalization power in one layer convolutional kernels.

关键词：

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：