Batch Normalization (BN) uses mini-batch statistics to normalize the
activations during training, introducing dependence between mini-batch
elements. This dependency can hurt the performance if the mini-batch size is
too small, or if the elements are correlated. Several alternatives, such as
Batch Renormalization and Group Normalization (GN), have been proposed to
address this issue. However, they either do not match the performance of BN for
large batches, or still exhibit degradation in performance for smaller batches,
or introduce artificial constraints on the model architecture. In this paper we
propose the Filter Response Normalization (FRN) layer, a novel combination of a
normalization and an activation function, that can be used as a replacement for
other normalizations and activations. Our method operates on each activation
channel of each batch element independently, eliminating the dependency on
other batch elements. Our method outperforms BN and other alternatives in a
variety of settings for all batch sizes. FRN layer performs $\approx 0.7-1.0\%$
better than BN on top-1 validation accuracy with large mini-batch sizes for
Imagenet classification using InceptionV3 and ResnetV2-50 architectures.
Further, it performs $>1\%$ better than GN on the same problem in the small
mini-batch size regime. For object detection problem on COCO dataset, FRN layer
outperforms all other methods by at least $0.3-0.5\%$ in all batch size
regimes.

本文提出了 Filter Response Normalization（FRN）层，一种归一化和激活函数的新型组合，可以作为其他归一化和激活函数的替代品；该方法在各种情况下都优于 BN 和其他替代品，对于具有大型 mini-batch 的 Imagenet 分类使用 InceptionV3 和 ResnetV2-50 架构，FRN 层比 BN 的 top-1 验证精度增加约 0.7-1.0％；对于小型 mini-batch，它比 GN 的性能增强超过 1％；对于 COCO 数据集上的目标检测问题，FRN 层在所有批处理大小各个方面都优于其他方法 0.3-0.5％。