Propagation of quantization error in performing intra-prediction with deep learning

Raz Birman, Yoram Segal, Avishay David-Malka, Ofer Hadar, Ron Shmueli

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review


Standard video compression algorithms use multiple "Modes", which are various linear combinations of pixels for prediction of their neighbors within image Macro-Blocks (MBs). In this research, we are using Deep Neural Networks (DNN) with supervised learning to predict block pixels. Using DNNs and employing intra-block pixel values' calculations that penetrate into the block, we manage to obtain improved predictions that yield up to 200% reduction of residual block errors. However, using intra-block pixels for predictions brings upon interesting tradeoffs between prediction errors and quantization errors. We explore and explain these tradeoffs for two different DNN types. We further discovered that it is possible to achieve a larger dynamic range of quantization parameter (Qp) and thus reach lower bit-rates than standard modes, which already saturate at these Qp levels. We explore this phenomenon and explain its reasoning.

Original languageEnglish
Title of host publicationApplications of Digital Image Processing XLII
EditorsAndrew G. Tescher, Touradj Ebrahimi
ISBN (Electronic)9781510629677
StatePublished - 2019
Externally publishedYes
EventApplications of Digital Image Processing XLII 2019 - San Diego, United States
Duration: 12 Aug 201915 Aug 2019

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
ISSN (Print)0277-786X
ISSN (Electronic)1996-756X


ConferenceApplications of Digital Image Processing XLII 2019
Country/TerritoryUnited States
CitySan Diego

Bibliographical note

Publisher Copyright:
© 2019 SPIE.


Dive into the research topics of 'Propagation of quantization error in performing intra-prediction with deep learning'. Together they form a unique fingerprint.

Cite this