Abstract

Mid-level visual element discovery aims to fifind clusters of image patches that are both representative and discriminative. In this work, we study this problem from the prospective of pattern mining while relying on the recently popularized Convolutional Neural Networks (CNNs). Specififically, we fifind that for an image patch, activation extracted from the fifirst fully-connected layer of a CNN have two appealing properties which enable its seamless integration with pattern mining. Patterns are then discovered from a large number of CNN activations of image patches through the wellknown association rule mining. When we retrieve and visualize image patches with the same pattern (See Fig. 1), surprisingly, they are not only visually similar but also semantically consistent. We apply our approach to scene and object classifification tasks, and demonstrate that our approach outperforms all previous works on mid-level visual element discovery by a sizeable margin with far fewer elements being used. Our approach also outperforms or matches recent works using CNN for these tasks. Source code of the complete system is available online.