Improved Residual Networks for Image and Video Recognition.pdf

一篇文献,基于视频的人体行为识别综述 A Review on Video-Based Human Activity Recognition

Identity Mappings in Deep Residual Networks Abstract 深度残差网络作为一种极深的网络结构,它展现了其极好的准确性和收敛行为。在本文中,我们分析了

kaiminghe论文：icml2016_tutorial_deep_residual_networks_kaiminghe

Recurrentneuralnetworks(RNNs)areapowerfulmodelforsequentialdata.End-to-endtrainingmethodssuchasConn

ImageCompressionforWirelessSensorNetworks

中文视频中字幕自动检测识别,张斌,张洪刚,视频中字幕文字的自动定位及识别已经成为视频标注及检索系统中重要的组成部分。本文提出了一个自动定位、识别视频中字幕文字的系

An improved FCM algorithm for ripe fruit image segmentation

We present a global optimization method, called the genetic algorithms (GAs), for digital image/spec

Developingneuralnetworkimageclassificationmodelsoftenrequiressignificantarchitectureengineering.Inth

系统为图像的缺陷,并设计了一种决策算法粗糙集适用于图像缺陷识别。首先,图像是由离散集区域化和顺序提出了连续属性离散化的图像。那么这个决策表模型离散的条件属性和决策属性的构造。进一步的函数和减少条件属性