Unsupervised Visual Representation Learning by Context Prediction

This work explores the use of spatial context as a source of free and plentiful supervisory signal for training a rich visual representation. Given only a large, unlabeled image collection, we extract random pairs of patches from each image and train a convolutional neural net to predict the position of the second patch relative to the first. We argue that doing well on this task requires the model to learn to recognize objects and their parts. We demonstrate that the feature representation learned using this within-image context indeed captures visual similarity across images. For example, this representation allows us to perform unsupervised visual discovery of objects like cats, people, and even birds from the Pascal VOC 2011 detection dataset. Furthermore, we show that the learned ConvNet can be used in the R-CNN framework [19] and provides a significant boost over a randomly-initialized ConvNet, resulting in state-of-the-art performance among algorithms which use only Pascal-provided training set annotations.

1. Introduction

Recently, new computer vision methods have leveraged large datasets of millions of labeled examples to learn rich, high-performance visual representations [29]. Yet efforts to scale these methods to truly Internet-scale datasets (i.e. hundreds of billions of images) are hampered by the sheer expense of the human annotation required. A natural way to address this difficulty would be to employ unsupervised learning, which aims to use data without any annotation. Unfortunately, despite several decades of sustained effort, unsupervised methods have not yet been shown to extract useful information from large collections of full-sized, real images. After all, without labels, it is not even clear what should be represented. How can one write an objective function to encourage a representation to capture, for example, objects, if none of the objects are labeled?