arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1611.05552cs.CVcs.LGcs.NE
收藏

DelugeNets:具有高效灵活的跨层信息流入的深度网络

DelugeNets: Deep Networks with Efficient and Flexible Cross-layer Information Inflows

Nanyang Technological University(南洋理工大学);Alibaba Group(阿里巴巴集团)

Jason Kuen, Xiangfei Kong, Gang Wang, Yap-Peng Tan

摘要:

神经网络是一种深度神经网络,它有效地促进了大量跨层信息从前一层流入到后一层。AdjugeNets中各层之间的连接是通过具有可学习过滤器的跨层深度卷积层建立的,作为一种灵活而高效的选择机制。与ResNets相比,MogeNets可以更灵活地跨多个层传播信息,并更有效地利用网络参数,同时比DenseNets更高效。值得注意的是,一个模型复杂度仅为4.31 GigaFLOPs和20.2M网络参数的AdjugeNet模型,在CIFAR-10和CIFAR-100数据集上分别实现了3.76%和19.02%的分类错误。此外,在ImageNet数据集上,ResNet-122的表现与ResNet-200相当,尽管其计算成本仅为后者的一半。

英文摘要:

Deluge Networks (DelugeNets) are deep neural networks which efficiently facilitate massive cross-layer information inflows from preceding layers to succeeding layers. The connections between layers in DelugeNets are established through cross-layer depthwise convolutional layers with learnable filters, acting as a flexible yet efficient selection mechanism. DelugeNets can propagate information across many layers with greater flexibility and utilize network parameters more effectively compared to ResNets, whilst being more efficient than DenseNets. Remarkably, a DelugeNet model with just model complexity of 4.31 GigaFLOPs and 20.2M network parameters, achieve classification errors of 3.76% and 19.02% on CIFAR-10 and CIFAR-100 dataset respectively. Moreover, DelugeNet-122 performs competitively to ResNet-200 on ImageNet dataset, despite costing merely half of the computations needed by the latter.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑