Copy-and-Paste Networks for Deep Video Inpainting

TL;DR

This paper introduces Copy-and-Paste Networks, a deep learning framework for video inpainting that leverages temporal information and frame alignment to produce coherent, fast, and visually pleasing results, also extending to exposure correction.

Contribution

The paper proposes a novel DNN-based framework for video inpainting that uses copying, pasting, and alignment mechanisms to improve temporal coherence and speed over existing methods.

Findings

Produces visually pleasing, temporally coherent inpainting results.

Runs faster than state-of-the-art optimization-based methods.

Improves lane detection accuracy through exposure enhancement.

Abstract

We present a novel deep learning based algorithm for video inpainting. Video inpainting is a process of completing corrupted or missing regions in videos. Video inpainting has additional challenges compared to image inpainting due to the extra temporal information as well as the need for maintaining the temporal coherency. We propose a novel DNN-based framework called the Copy-and-Paste Networks for video inpainting that takes advantage of additional information in other frames of the video. The network is trained to copy corresponding contents in reference frames and paste them to fill the holes in the target frame. Our network also includes an alignment network that computes affine matrices between frames for the alignment, enabling the network to take information from more distant frames for robustness. Our method produces visually pleasing and temporally coherent results while running faster than the state-of-the-art optimization-based method. In addition, we extend our framework for enhancing over/under exposed frames in videos. Using this enhancement technique, we were able to significantly improve the lane detection accuracy on road videos.