Dynamic ReLU
arxiv(2020)
摘要
Rectified linear units (ReLU) are commonly used in deep neural networks. So far ReLU and its generalizations (either non-parametric or parametric) are static, performing identically for all input samples. In this paper, we propose Dynamic ReLU (DY-ReLU), a dynamic rectifier whose parameters are input-dependent as a hyper function over all input elements. The key insight is that DY-ReLU encodes the global context into the hyper function and adapts the piecewise linear activation function accordingly. Compared to its static counterpart, DY-ReLU has negligible extra computational cost, but significantly more representation capability, especially for light-weight neural networks. By simply using DY-ReLU for MobileNetV2, the top-1 accuracy on ImageNet classification is boosted from 72.0% to 76.2% with only 5% additional FLOPs.
更多查看译文
关键词
ReLU, Convolutional Neural Networks, Dynamic
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络