English

Learning a Discriminative Model for the Perception of Realism in Composite Images

Computer Vision and Pattern Recognition 2015-10-05 v1

Abstract

What makes an image appear realistic? In this work, we are answering this question from a data-driven perspective by learning the perception of visual realism directly from large amounts of data. In particular, we train a Convolutional Neural Network (CNN) model that distinguishes natural photographs from automatically generated composite images. The model learns to predict visual realism of a scene in terms of color, lighting and texture compatibility, without any human annotations pertaining to it. Our model outperforms previous works that rely on hand-crafted heuristics, for the task of classifying realistic vs. unrealistic photos. Furthermore, we apply our learned model to compute optimal parameters of a compositing method, to maximize the visual realism score predicted by our CNN model. We demonstrate its advantage against existing methods via a human perception study.

Keywords

Cite

@article{arxiv.1510.00477,
  title  = {Learning a Discriminative Model for the Perception of Realism in Composite Images},
  author = {Jun-Yan Zhu and Philipp Krähenbühl and Eli Shechtman and Alexei A. Efros},
  journal= {arXiv preprint arXiv:1510.00477},
  year   = {2015}
}

Comments

International Conference on Computer Vision (ICCV) 2015