Decision Trees for Complexity Reduction in Video Compression

  • 2019-08-12 14:27:16
  • Natasha Westland, AndrĂ© Seixas Dias, Marta Mrak
  • 1

Abstract

This paper proposes a method for complexity reduction in practical videoencoders using multiple decision tree classifiers. The method is demonstratedfor the fast implementation of the 'High Efficiency Video Coding' (HEVC)standard, chosen because of its high bit rate reduction capability but largecomplexity overhead. Optimal partitioning of each video frame into coding units(CUs) is the main source of complexity as a vast number of combinations aretested. The decision tree models were trained to identify when the CU testingprocess, a time-consuming Lagrangian optimisation, can be skipped i.e a highprobability that the CU can remain whole. A novel approach to finding thesimplest and most effective decision tree model called 'manual pruning' isdescribed. Implementing the skip criteria reduced the average encoding time by42.1% for a Bj{\o}ntegaard Delta rate detriment of 0.7%, for 17 standard testsequences in a range of resolutions and quantisation parameters.

 

Quick Read (beta)

This feature is not avaialbe for this paper.