计算机科学
推论
编码(集合论)
学习迁移
卷积神经网络
人工智能
比例(比率)
分辨率(逻辑)
人工神经网络
算法
缩放比例
机器学习
模式识别(心理学)
数学
物理
集合(抽象数据类型)
程序设计语言
量子力学
几何学
作者
Mingxing Tan,Quoc V. Le
出处
期刊:International Conference on Machine Learning
日期:2019-05-24
卷期号:: 6105-6114
被引量:1324
摘要
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 84.3% top-1 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.
科研通智能强力驱动
Strongly Powered by AbleSci AI