Context Encoders: Feature Learning by Inpainting

TLDR

The authors introduce Context Encoders, an unsupervised visual feature learning algorithm that predicts missing image regions from surrounding context. These convolutional neural networks are trained with pixel‑wise reconstruction or reconstruction plus adversarial loss, requiring global image understanding to generate plausible hypotheses for the missing parts. Adversarial training produces sharper results, and the learned representation improves CNN pre‑training for classification, detection, and segmentation, while also enabling effective semantic inpainting.

Abstract

We present an unsupervised visual feature learning algorithm driven by context-based pixel prediction. By analogy with auto-encoders, we propose Context Encoders - a convolutional neural network trained to generate the contents of an arbitrary image region conditioned on its surroundings. In order to succeed at this task, context encoders need to both understand the content of the entire image, as well as produce a plausible hypothesis for the missing part(s). When training context encoders, we have experimented with both a standard pixel-wise reconstruction loss, as well as a reconstruction plus an adversarial loss. The latter produces much sharper results because it can better handle multiple modes in the output. We found that a context encoder learns a representation that captures not just appearance but also the semantics of visual structures. We quantitatively demonstrate the effectiveness of our learned features for CNN pre-training on classification, detection, and segmentation tasks. Furthermore, context encoders can be used for semantic inpainting tasks, either stand-alone or as initialization for non-parametric methods.