具有密集连接性的卷积网络（cs Machine learning）
- 2020 年 1 月 10 日
- 笔记
最近的研究表明，如果卷积网络在靠近输入的层和靠近输出的层之间包含较短的连接，则可以进行更深入，更准确和有效的训练。在本文中，我们接受了这一说法，并介绍了密集卷积网络（DenseNet)。该网络以前馈的方式将每一层连接到其他每一层，层次结构为具有L层以及L连接的传统卷积网络-每个层与其后一层之间有一个连接-我们的网络具有L（L + 1）/ 2个直接连接。对于每个层，前面所有层的功能映射都可以输入，而它自己的功能映射可以使用所有后续层的输入。DenseNets具有几个引人注目的优势：它们减轻了梯度消失的问题，鼓励功能复用，并大幅提高参数效率。我们在四个竞争激烈的对象识别基准测试任务（CIFAR-10，CIFAR-100，SVHN和ImageNet）上评估了我们提出的体系结构。DenseNets在大多数方面都获得了超越现有技术的显着改进，同时需要更少的参数和计算即可实现高性能。
原文题目：Convolutional Networks with Dense Connectivity
原文：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, encourage feature reuse and substantially improve parameter efficiency. 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 parameters and computation to achieve high performance.
原文作者：Gao Huang,Zhuang Liu,Geoff Pleiss,Laurens van der Maaten,Kilian Q. Weinberger
原文地址：https://arxiv.org/abs/2001.02394
