Unsupervised Representation Learning by Predicting Image Rotations


Nov 03, 2017 (modified: Nov 03, 2017) ICLR 2018 Conference Blind Submission readers: everyone Show Bibtex
  • Abstract: Over the last years, deep convolutional neural networks (ConvNets) have transformed the field of computer vision thanks to their unparalleled capacity to learn high level semantic image features. However, in order to successfully learn those features, they usually require massive amounts of manually labeled data, which is both expensive and impractical to scale. Therefore, unsupervised semantic feature learning, i.e., learning without requiring manual annotation effort, is of crucial importance in order to successfully harvest the vast amount of visual data that are available today. In our work we propose to learn image features by training ConvNets to recognize the 2d rotation that is applied to the image that it gets as input. We demonstrate both qualitative and quantitative that this apparently simple task actually provides a very powerful supervisory signal for semantic feature learning. We exhaustively evaluate our method in various unsupervised feature learning benchmarks and we exhibit in all of them remarkably good performance. Even more, our results on those benchmarks demonstrate dramatic improvements w.r.t. prior state-of-the-art, thus they significantly close the gap between unsupervised and supervised feature learning. For instance, in PASCAL VOC 2007 detection task our unsupervised pre-trained AlexNet model achieves the state-of-the-art (among unsupervised methods) mAP of $54.4\%$ that is only 2.4 points lower from the supervised case. Similar striking results we get when transfer our unsupervised learned features on various other tasks, such as ImageNet classification, PASCAL classification, and CIFAR-10 classification.
  • Keywords: Unsupervised representation learning