在模型扩展时平衡好深度、宽度、分辨率,取得精度、效率、模型大小的最大化。借由此简单有效的模型扩展方法,作者在使用神经架构搜索得到的基模型上扩展出一系列EfficientNets模型,达到了更好的精度和效率的平衡。该文已被ICML 2019录用,这可能是一篇要改变整个深度卷积网络模型设计的论文了。
Convolutional Neural Networks (ConvNets) are commonly developed at a fixed resource budget, and then scaled up for better accuracy if more resources are available. In this paper, we systematically study model scaling and identify that carefully balancing network depth, width, and resolution can lead to better performance. Based on this observation, we propose a new scaling method that uniformly scales all dimensions of depth/width/resolution using a simple yet highly effective compound coefficient. We demonstrate the effectiveness of this method on scaling up MobileNets and ResNet.
To go even further, we use neural architecture search to design a new baseline network and scale it up to obtain a family of models, called EfficientNets, which achieve much better accuracy and efficiency than previous ConvNets. In particular, our EfficientNet-B7 achieves state-of-the-art 84.4% top-1 / 97.1% top-5 accuracy on ImageNet, while being 8.4x smaller and 6.1x faster on inference than the best existing ConvNet. Our EfficientNets also transfer well and achieve state-of-the-art accuracy on CIFAR-100 (91.7%), Flowers (98.8%), and 3 other transfer learning datasets, with an order of magnitude fewer parameters. Source code is at this https URL.
卷积神经网络(ConvNets)通常是在固定资源预算下开发的,如果有更多可用资源,则可以按比例放大以获得更高的准确性。在本文中,我们系统地研究模型缩放并确定仔细平衡网络深度,宽度和分辨率可以带来更好的性能。基于这一观察,我们提出了一种新的缩放方法,该方法使用简单但高效的复合系数均匀地缩放深度/宽度/分辨率的所有尺寸。我们证明了这种方法在扩展MobileNets和ResNet方面的有效性。为了更进一步,我们使用神经架构搜索来设计新的基线网络并进行扩展以获得一系列模型,称为EfficientNets,它比以前的ConvNets具有更高的准确性和效率。特别是,我们的EfficientNet-B7在ImageNet上实现了最先进的84.4%前1 / 97.1%前5精度,同时比最好的现有ConvNet小了8.4倍,推理速度快6.1倍。我们的EfficientNets在CIFAR-100(91.7%),Flowers(98.8%)和其他3个传输学习数据集上也能很好地传输和实现最先进的精度,参数的数量级减少了一个数量级。源代码位于此https URL。