DropBlock: A regularization method for convolutional networks

TL;DR

DropBlock is a structured dropout method that drops contiguous regions in feature maps, improving regularization and accuracy of convolutional neural networks on tasks like image classification and object detection.

Contribution

This paper introduces DropBlock, a novel structured dropout technique for convolutional layers that enhances regularization and improves performance over traditional dropout.

Findings

DropBlock outperforms dropout in regularizing CNNs.

ResNet-50 with DropBlock achieves 78.13% on ImageNet.

DropBlock increases COCO detection AP from 36.8% to 38.4%.

Abstract

Deep neural networks often work well when they are over-parameterized and trained with a massive amount of noise and regularization, such as weight decay and dropout. Although dropout is widely used as a regularization technique for fully connected layers, it is often less effective for convolutional layers. This lack of success of dropout for convolutional layers is perhaps due to the fact that activation units in convolutional layers are spatially correlated so information can still flow through convolutional networks despite dropout. Thus a structured form of dropout is needed to regularize convolutional networks. In this paper, we introduce DropBlock, a form of structured dropout, where units in a contiguous region of a feature map are dropped together. We found that applying DropbBlock in skip connections in addition to the convolution layers increases the accuracy. Also, gradually increasing number of dropped units during training leads to better accuracy and more robust to hyperparameter choices. Extensive experiments show that DropBlock works better than dropout in regularizing convolutional networks. On ImageNet classification, ResNet-50 architecture with DropBlock achieves $78.13\%$ accuracy, which is more than $1.6\%$ improvement on the baseline. On COCO detection, DropBlock improves Average Precision of RetinaNet from $36.8\%$ to $38.4\%$.