Discriminative Unsupervised Feature Learning with Exemplar Convolutional Neural Networks
Alexey Dosovitskiy, Philipp Fischer, Jost Tobias Springenberg, Martin Riedmiller, Thomas Brox All authors are with the Computer Science Department at the University of Freiburg E-mail: {dosovits, fischer, springj, riedmiller,
Abstract
Deep convolutional networks have proven to be very successful in learning task specific features that allow for unprecedented performance on various computer vision tasks. Training of such networks follows mostly the supervised learning paradigm, where sufficiently many input-output pairs are required for training. Acquisition of large training sets is one of the key challenges, when approaching a new task. In this paper, we aim for generic feature learning and present an approach for training a convolutional network using only unlabeled data. To this end, we train the network to discriminate between a set of surrogate classes. Each surrogate class is formed by applying a variety of transformations to a randomly sampled ’seed’ image patch. In contrast to supervised network training, the resulting feature representation is not class specific. It rather provides robustness to the transformations that have been applied during training. This generic feature representation allows for classification results that outperform the state of the art for unsupervised learning on several popular datasets (STL-10, CIFAR-10, Caltech-101, Caltech-256). While such generic features cannot compete wit
中文速览
监督学习卷积网络需要海量标注数据，这一瓶颈在很多场景下难以克服，而无监督特征学习又长期在性能上远落后于有监督方法。本文提出了一种叫做 Exemplar-CNN 的无监督卷积网络训练方法：从无标注图像中随机采样大量图像块作为"种子"，对每个种子施加平移、缩放、旋转、色彩变换等多种随机变换，将同一种子生成的所有变换版本定义为一个"代理类别（surrogate class）"，然后训练网络区分这数万个代理类别。这样训练出的网络自然学会了对上述变换保持不变性，同时又能区分不同的图像内容，获得一种通用的图像特征表示。实验表明，该方法在 STL-10、CIFAR-10、Caltech-101、Caltech-256 四个数据集上的无监督分类精度超过了此前所有无监督方法，在几何匹配任务上甚至超过了用 ImageNet 有监督训练的 AlexNet 特征和经典的 SIFT 描述子，证明了完全不依赖标注数据也能学到强泛化能力特征的可行性。
原文 arXiv:1406.6909;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1406.6909v2