水准点(测量)
特征(语言学)
计算机科学
图层(电子)
卷积神经网络
计算
重新使用
编码(集合论)
对象(语法)
网络体系结构
模式识别(心理学)
人工智能
算法
计算机网络
集合(抽象数据类型)
程序设计语言
工程类
废物管理
有机化学
地理
化学
哲学
语言学
大地测量学
作者
Gao Huang,Zhuang Liu,Laurens van der Maaten,Kilian Q. Weinberger
出处
期刊:Cornell University - arXiv
日期:2016-08-24
被引量:648
标识
DOI:10.48550/arxiv.1608.06993
摘要
Recent work has shown that convolutional networks can be substantially deeper, more accurate, and efficient to train if they contain shorter connections between layers close to the input and those close to the output. In this paper, we embrace this observation and introduce the Dense Convolutional Network (DenseNet), which connects each layer to every other layer in a feed-forward fashion. Whereas traditional convolutional networks with L layers have L connections - one between each layer and its subsequent layer - our network has L(L+1)/2 direct connections. For each layer, the feature-maps of all preceding layers are used as inputs, and its own feature-maps are used as inputs into all subsequent layers. DenseNets have several compelling advantages: they alleviate the vanishing-gradient problem, strengthen feature propagation, encourage feature reuse, and substantially reduce the number of parameters. We evaluate our proposed architecture on four highly competitive object recognition benchmark tasks (CIFAR-10, CIFAR-100, SVHN, and ImageNet). DenseNets obtain significant improvements over the state-of-the-art on most of them, whilst requiring less computation to achieve high performance. Code and pre-trained models are available at https://github.com/liuzhuang13/DenseNet .
科研通智能强力驱动
Strongly Powered by AbleSci AI