Learning Deep Features for Discriminative Localization

In this work, we revisit the global average pooling layer proposed in [13], and shed light on how it explicitly enables the convolutional neural network (CNN) to have remarkable localization ability despite being trained on imagelevel labels. While this technique was previously proposed as a means for regularizing training, we find that it actually builds a generic localizable deep representation that exposes the implicit attention of CNNs on an image. Despite the apparent simplicity of global average pooling, we are able to achieve 37.1% top-5 error for object localization on ILSVRC 2014 without training on any bounding box annotation. We demonstrate in a variety of experiments that our network is able to localize the discriminative image regions despite just being trained for solving classification task1.

1. Introduction

Recent work by Zhou et al [34] has shown that the convolutional units of various layers of convolutional neural networks (CNNs) actually behave as object detectors despite no supervision on the location of the object was provided. Despite having this remarkable ability to localize objects in the convolutional layers, this ability is lost when fully-connected layers are used for classification. Recently some popular fully-convolutional neural networks such as the Network in Network (NIN) [13] and GoogLeNet [25] have been proposed to avoid the use of fully-connected layers to minimize the number of parameters while maintaining high performance.