Forward and Backward Warping for Optical Flow-Based Frame Interpolation

Joi Shimizu, Heming Sun*, Jiro Katto

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Frame interpolation methods generate intermediate frames by taking consecutive frames as inputs. This enables the generation of high frame rate videos from low frame rate videos. Recently, many deep learning-based frame interpolation methods have been proposed. One way of frame interpolation is by using the bi-directional optical flow. In many cases, these methods use backward warping to warp the input images to the desired frame. However, forward warping can also be used to warp the input frames. In this paper, we propose a frame interpolation method that utilizes both forward warping and backward warping. Experimental results show that utilizing both warping methods can enhance the performance compared to only using backward warping.

Original languageEnglish
Title of host publication4th International Conference on Artificial Intelligence in Information and Communication, ICAIIC 2022 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages82-86
Number of pages5
ISBN (Electronic)9781665458184
DOIs
Publication statusPublished - 2022
Event4th International Conference on Artificial Intelligence in Information and Communication, ICAIIC 2022 - Jeju lsland, Korea, Republic of
Duration: 2022 Feb 212022 Feb 24

Publication series

Name4th International Conference on Artificial Intelligence in Information and Communication, ICAIIC 2022 - Proceedings

Conference

Conference4th International Conference on Artificial Intelligence in Information and Communication, ICAIIC 2022
Country/TerritoryKorea, Republic of
CityJeju lsland
Period22/2/2122/2/24

Keywords

  • deep learning
  • frame interpolation
  • optical flow

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Science Applications
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Forward and Backward Warping for Optical Flow-Based Frame Interpolation'. Together they form a unique fingerprint.

Cite this