Reading: Jin VCIP’17 — CNN Oriented Fast QTBT Parititon (Fast VVC Prediction)

Original article was published on Artificial Intelligence on Medium

Reading: Jin VCIP’17 — CNN Oriented Fast QTBT Parititon (Fast VVC Prediction)

42.80% Complexity Reduction With Only 0.65% Increase in BD-Rate

In this story, CNN Oriented Fast QTBT Partition Algorithm for JVET Intra Coding (Jin VCIP’17), by Shanghai University, is presented. I read this because I work on video coding research. In this paper, a fast encoding approach for quad-tree plus binary-tree (QTBT) is proposed in Versatile Video Coding (VVC):

  • The QTBT partition depth range is modelled as a multi-class classification problem.
  • The depth range of 32×32 block is predicted directly, rather than to judge split or not at each depth level.

This is a paper in 2017 VCIP. (Sik-Ho Tsang @ Medium)