Fast Transform Decision Scheme for VVC Intra-Frame Prediction Using Decision Trees
Mário Saldanha, Gustavo Sanchez, César Marcon, Luciano Agostini
Abstract
This paper presents a fast transform decision scheme for Versatile Video Coding (VVC) intra-frame prediction using decision trees. VVC introduces several novel coding tools to improve the coding efficiency of the intra-frame prediction at the cost of a high computational effort, including a new transform coding process using Multiple Transform Selection (MTS) for primary transform and Low-Frequency Non-Separable Transform (LFNST) for secondary transform. We developed an efficient complexity reduction scheme composed of two solutions based on decision tree classifiers to avoid the MTS and LFNST evaluations in the costly Rate-Distortion optimization (RDO) process. Experimental results showed that the proposed scheme provides 11% of encoding timesaving with a negligible impact on the coding efficiency.