English

SaiNet: Stereo aware inpainting behind objects with generative networks

Computer Vision and Pattern Recognition 2022-05-17 v1 Graphics Machine Learning

Abstract

In this work, we present an end-to-end network for stereo-consistent image inpainting with the objective of inpainting large missing regions behind objects. The proposed model consists of an edge-guided UNet-like network using Partial Convolutions. We enforce multi-view stereo consistency by introducing a disparity loss. More importantly, we develop a training scheme where the model is learned from realistic stereo masks representing object occlusions, instead of the more common random masks. The technique is trained in a supervised way. Our evaluation shows competitive results compared to previous state-of-the-art techniques.

Keywords

Cite

@article{arxiv.2205.07014,
  title  = {SaiNet: Stereo aware inpainting behind objects with generative networks},
  author = {Violeta Menéndez González and Andrew Gilbert and Graeme Phillipson and Stephen Jolly and Simon Hadfield},
  journal= {arXiv preprint arXiv:2205.07014},
  year   = {2022}
}

Comments

Presented at AI4CC workshop at CVPR

R2 v1 2026-06-24T11:17:15.047Z