Neural weight step video compression

Czerkawski, Mikolaj and Cardona, Javier and Atkinson, Robert and Michie, Craig and Andonovic, Ivan and Clemente, Carmine and Tachtatzis, Christos (2021) Neural weight step video compression. In: NeurIPS 2021 Pre-Registration Workshop, 2021-12-13 - 2021-12-13, Online.

Preview	Text. Filename: Czerkawski_etal_PreReg_2021_Neural_weight_step_video_compression.pdf Final Published Version License: Download (172kB)| Preview
	Video. Filename: Czerkawski_etal_PreReg_2021_Neural_weight_step_video_compression_video_summary.mp4 License: Download (7MB)

Abstract

A variety of compression methods based on encoding images as weights of a neural network have been recently proposed. Yet, the potential of similar approaches for video compression remains unexplored. In this work, we suggest a set of experiments for testing the feasibility of compressing video using two architectural paradigms, coordinate-based MLP (CbMLP) and convolutional network. Furthermore, we propose a novel technique of neural weight stepping, where subsequent frames of a video are encoded as low-entropy parameter updates. To assess the feasibility of the considered approaches, we will test the video compression performance on several high-resolution video datasets and compare against existing conventional and neural compression techniques.

ORCID iDs

Czerkawski, Mikolaj ORCID: https://orcid.org/0000-0002-0927-0416

Cardona, Javier ORCID: https://orcid.org/0000-0002-9284-1899

Atkinson, Robert ORCID: https://orcid.org/0000-0002-6206-2229

Michie, Craig ORCID: https://orcid.org/0000-0001-5132-4572

Andonovic, Ivan ORCID: https://orcid.org/0000-0001-9093-5245

Clemente, Carmine ORCID: https://orcid.org/0000-0002-6665-693X

Tachtatzis, Christos ORCID: https://orcid.org/0000-0001-9150-6805

• Share and Export
  

  RefWorksRIS (EndNote Online, Zotero, Ref manager, etc)RDF+XMLEndNoteASCII CitationOpenURL ContextObject in SpanMPEG-21 DIDLRIOXX2 XMLBibTeXAtomEP3 XMLDublin CoreRDF+N-TriplesMODSHTML CitationJSONData Cite XMLRDF+N3Multiline CSVOpenURL ContextObjectReferSimple MetadataMETS
• Item metadata

  Item type:	Conference or Workshop Item(Paper)
  ID code:	78890
  Dates:	Date Event 13 December 2021 Published 21 October 2021 Accepted
  Subjects:	Technology > Electrical engineering. Electronics Nuclear engineering
  Department:	Faculty of Engineering > Electronic and Electrical Engineering Faculty of Engineering > Chemical and Process Engineering Strategic Research Themes > Measurement Science and Enabling Technologies
  Depositing user:	Pure Administrator
  Date deposited:	13 Dec 2021 14:34
  Last modified:	11 Nov 2024 17:05
  URI:	https://strathprints.strath.ac.uk/id/eprint/78890