Back to Search Document details
16th Meeting: Geneva, October 2019 2019-09-17 10:56
CE6-1: LFNST with one mode
Abstract
This document provides results for core experiment CE6-1 which was set up to study effectiveness of different simplifications to LFNST transform selection process. It is asserted that the results would justify adopting the normative change of either CE6-1.1a or CE6-1.1d as both provide around 10 % encoder runtime decrease for around 0.1 % BD-rate impact in the AI configuration.
JVET-P0041 CE6-1: LFNST with one mode [M.-S. Chiang, C.-W. Hsu, Y.-W. Huang, S.-M. Lei (MediaTek), J. Lainema (Nokia)]

This document provides results for core experiment CE6-1 which was set up to study effectiveness of different simplifications to LFNST transform selection process. It is asserted that the results would justify adopting the normative change of either CE6-1.1a or CE6-1.1d as both provide around 10 % encoder runtime decrease for around 0.1 % BD-rate impact in the AI configuration.

PATENTS:
US12114015B2 0.71 2024-10-08 US12231685B2 0.71 2025-02-18 US11991393B2 0.32 2024-05-21 US11943476B2 0.30 2024-03-26 TW202337207A 0.28 2023-09-16 US11616966B2 0.26 2023-03-28 EP3866468B1 0.24 2025-11-19 US20250063155A1 0.20 2025-02-20
Decisions
adopted
This document provides results for core experiment CE6-1 which was set up to study effectiveness of different simplifications to LFNST transform selection process. It is asserted that the results would justify adopting the normative change of either CE6-1.1a or CE6-1.1d as both provide around 10 % encoder runtime decrease for around 0.1 % BD-rate impact in the AI configuration.
Citation