Bayesian adaptive algorithm for fast coding unit decision in the High Efficiency Video Coding (HEVC) standard

作者：

Highlights：

• An effective early CU depth decision algorithm to reduce the encoder complexity is proposed.

• A hypothesis test based on Bayesian statistics is designed to make a decision at every CU depth in a hierarchical way.

• The statistical parameters are updated on the fly to the changing content of the sequence achieving an adaptive design.

• R-D costs are employed as features to make accurate early decisions on the CU depth.

摘要

Highlights•An effective early CU depth decision algorithm to reduce the encoder complexity is proposed.•A hypothesis test based on Bayesian statistics is designed to make a decision at every CU depth in a hierarchical way.•The statistical parameters are updated on the fly to the changing content of the sequence achieving an adaptive design.•R-D costs are employed as features to make accurate early decisions on the CU depth.

论文关键词：Bayesian statistics,Complexity reduction,Fast coding unit decision,High Efficiency Video Coding,Hypothesis test,On the fly estimation

论文评审过程：Received 5 July 2016, Revised 4 April 2017, Accepted 5 April 2017, Available online 8 April 2017, Version of Record 25 April 2017.

论文官网地址：https://doi.org/10.1016/j.image.2017.04.004