Organisation internationale de normalisation

CE1 primary contributions (7)

1.1.1.1.1.1.1.1.38JCTVC-U0050 CE1 Test B: Simplification for index map coding in palette mode [J. Kim, J. Ye, P. Lai, S. Liu, S. Lei (MediaTek)]
1.1.1.1.1.1.1.1.39JCTVC-U0061 CE1: Test A.1: Extended copy above mode to the first line with index adjustment bits [Y.-C. Sun, J. Kim, T.-D. Chuang, S. Liu, Y.-W. Chen, Y.-W. Huang, S. Lei (MediaTek), V. Seregin, F. Zou, W. Pu, R. Joshi, M. Karczewicz (Qualcomm)]
1.1.1.1.1.1.1.1.40JCTVC-U0062 CE1: Test A.2: Extended copy above mode to the first line with grouping the palette_run_type_flag in front [Y.-C. Sun, J. Kim, T.-D. Chuang, S. Liu, Y.-W. Chen, Y.-W. Huang, S. Lei (MediaTek)]
1.1.1.1.1.1.1.1.41JCTVC-U0088 CE1 Test C.2: Palette run-length coding with rotated index scans (scan from bottom-right) [P. Lai, J. Kim, S. Liu, S. Lei (MediaTek)]
1.1.1.1.1.1.1.1.42JCTVC-U0101 CE1 Test C.1: Improved palette run-length coding with palette reverse scan order [X. Xiu, Y. Ye, Y. He (InterDigital), W. Pu, R. Joshi, M. Karczewicz, F. Zou, V. Seregin (Qualcomm)]
1.1.1.1.1.1.1.1.43JCTVC-U0122 CE1 Test D: Modification of Palette Run Coding [M. Karczewicz, W. Pu, R. Joshi, F. Zou, V. Seregin (Qualcomm)] [late]
1.1.1.1.1.1.1.1.44JCTVC-U0127 CE1: Additional results for CE1 subtest C [R. Joshi, M. Karczewicz, V. Seregin, W. Pu, F. Zou (Qualcomm)] [late]

(Consideration of this topic was chaired by GJS on Friday 06-19, 15:00-15:30.)

This was not originally planned in the CE, and was classified as information document.

Additional results for CE1 subtest C were provided to understand how much do the different aspects (encoder, scans) contribute to the overall gain. The scans without any encoder modifications to the way in which runs are calculated reportedly show BD-rate improvements in the range of 0.0% to 0.5% for both full frame IBC and 1×4 CTU IBC, All-Intra configurations. The BD-rate gains from encoder modifications to the way in which runs are calculated are in the range of 0.0% to 0.5% for full frame IBC and 0.0% to 0.6% for 1×4 CTU IBC, All-Intra configurations.

Thus, the gains shown in CE1 subtest C were shown to be a combination of two things:

• 
  Additional scan patterns (normative)
  
• 
  Improved run-type flag decisions and run value decisions (non-normative)
  

The gain from each of these sources is roughly half of the total gain shown.

See the notes on U0096, which includes a simplified form of the technique described in this contribution, and also includes some other techniques.

3.