Dynamic Style Adaptation Network: A Comprehensive Approach for Video Style Transfer

DOI

10.2991/978-94-6463-300-9_81How to use a DOI?

Keywords

Video style transfer; dynamic style adaption; deep learning

Abstract

Video style transfer is an emerging research hotspot in the computer vision community, which aims to apply artistic styles to videos to generate visually appealing and stylized video sequences. Compared to image style transfer, video style transfer mainly involves adjusting temporal consistency, which requires seamless style transfer across frames while maintaining temporal coherence. Thanks to the rapid development of Convolutional neural network, the accuracy and speed of video image migration have made breakthroughs, but there are still many challenges in balancing style fidelity, computing efficiency, and retaining the original video content. To alleviate the above issues, this paper proposes a video style consistency transfer algorithm (SCTAda) based on residual modules and adaptive attention. Specifically, SCTAda introduces residual modules to preserve the original content features, which is beneficial for improving the details of generated content. Then, SCTAda further introduces an adaptive attention module to selectively emphasize relevant style patterns in style videos and apply them to corresponding content frames, which helps improve coherence and accuracy. Extensive experiments have quantitatively and qualitatively verified the effectiveness of the method proposed in this paper, which indicate that SCTAda can generate high-quality videos with realistic content representation, coherent style patterns, and enhanced artistic quality in Video style transformation tasks.

Copyright

© 2023 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Ni Liu
PY  - 2023
DA  - 2023/11/27
TI  - Dynamic Style Adaptation Network: A Comprehensive Approach for Video Style Transfer
BT  - Proceedings of the 2023 International Conference on Image, Algorithms and Artificial Intelligence (ICIAAI 2023)
PB  - Atlantis Press
SP  - 776
EP  - 787
SN  - 2352-538X
UR  - https://doi.org/10.2991/978-94-6463-300-9_81
DO  - 10.2991/978-94-6463-300-9_81
ID  - Liu2023
ER  -