Neural Residual Flow Fields for Efficient Video Representations
- Authors
- Rho, D.[Rho, D.]; Cho, J.[Cho, J.]; Ko, J.H.[Ko, J.H.]; Park, E.[Park, E.]
- Issue Date
- 1-Jan-2023
- Publisher
- Springer Science and Business Media Deutschland GmbH
- Citation
- Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), v.13842 LNCS, pp.458 - 474
- Indexed
- SCOPUS
- Journal Title
- Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
- Volume
- 13842 LNCS
- Start Page
- 458
- End Page
- 474
- URI
- https://scholarworks.bwise.kr/skku/handle/2021.sw.skku/103857
- DOI
- 10.1007/978-3-031-26284-5_28
- ISSN
- 0302-9743
- Abstract
- Neural fields have emerged as a powerful paradigm for representing various signals, including videos. However, research on improving the parameter efficiency of neural fields is still in its early stages. Even though neural fields that map coordinates to colors can be used to encode video signals, this scheme does not exploit the spatial and temporal redundancy of video signals. Inspired by standard video compression algorithms, we propose a neural field architecture for representing and compressing videos that deliberately removes data redundancy through the use of motion information across video frames. Maintaining motion information, which is typically smoother and less complex than color signals, requires a far fewer number of parameters. Furthermore, reusing color values through motion information further improves the network parameter efficiency. In addition, we suggest using more than one reference frame for video frame reconstruction and separate networks, one for optical flows and the other for residuals. Experimental results have shown that the proposed method outperforms the baseline methods by a significant margin. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.
- Files in This Item
- There are no files associated with this item.
- Appears in
Collections - Information and Communication Engineering > School of Electronic and Electrical Engineering > 1. Journal Articles
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.