US 12,219,143 B2
Probability estimation for video coding
Jingning Han, Santa Clara, CA (US); Yue Sun, Mountain View, CA (US); and Yaowu Xu, Saratoga, CA (US)
Assigned to GOOGLE LLC, Mountain View, CA (US)
Appl. No. 17/775,565
Filed by Google LLC, Mountain View, CA (US)
PCT Filed Nov. 9, 2020, PCT No. PCT/US2020/059594
§ 371(c)(1), (2) Date May 9, 2022,
PCT Pub. No. WO2021/092531, PCT Pub. Date May 14, 2021.
Claims priority of provisional application 62/932,508, filed on Nov. 8, 2019.
Prior Publication US 2023/0007260 A1, Jan. 5, 2023
Int. Cl. H04N 19/13 (2014.01); H04N 19/126 (2014.01); H04N 19/91 (2014.01)
CPC H04N 19/13 (2014.11) [H04N 19/126 (2014.11); H04N 19/91 (2014.11)] 20 Claims
OG exemplary drawing
 
1. A method for entropy coding a sequence of symbols, comprising:
determining a first probability model for entropy coding the sequence, the first probability model being one of a plurality of available probability models;
entropy coding at least one symbol of the sequence using a probability determined by the first probability model;
after entropy coding the at least one symbol of the sequence, determining a first probability estimation to update the probability using the first probability model;
determining a second probability estimation to update the probability using a second probability model, wherein the second probability model is a different model to the first probability model; and
entropy coding a subsequent symbol relative to the at least one symbol of the sequence using the probability updated by a combination of the first probability estimation and the second probability estimation.