(See Fig.1~4)
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
FootballBusForemanMobile
ctxIdxInc2,flagOff
ctxIdxInc1,flagOff
ctxIdxInc0,flagOff
ctxIdxInc2,flagOn
ctxIdxInc1,flagOn
ctxIdxInc0,flagOn
Fig.1

**Probability**distribution at each context 0% 20% 40% 60% 80% 100% FootballBusForemanMobile ctxIdxInc0,flagOff ctxIdxInc0,flagOn Fig.2**Probability**distribution when ctxIdxInc = 0 0% 20% 40% 60% 80% 100% FootballBusForemanMobile ctxIdxInc1,flagOff ctxIdxInc1,flagOn Fig.3**Probability**distribution when ctxIdxInc = 1 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% FootballBusForemanMobile ctxIdxInc2,flagOff ctxIdxInc2,flagOn Fig.4**Probability**distribution when ctxIdxInc = 2 Fig.1 gives the whole distribution of three contexts. (...) If the base macroblock is coded as an intra mode, the**probability**of the base_mode_flag to be set increases up to 90% (refer to Fig.6) and in the opposite case that the base macroblock is coded as inter mode, the**probability**of the base_mode_flag to be set is about 55% (refer to Fig.7). The following figures show that the**probability**of base_mode_flag depends on MB type of base macroblock. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% BusFootballForemanMobile flagOff,baseInter flagOff,baseIntra flagOn,baseInter flagOn,baseIntra Fig.5**Probability**distribution according to MB type of base macroblock 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% BusFootballForemanMobile flagOff,baseIntra flagOn,baseIntra Fig.6**Probability**distribution when MB type of base macroblock is intra 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% BusFootballForemanMobile flagOff,baseInter flagOn,baseInter Fig.7**Probability**distribution when MB type of base macroblock is inter To clarify the effect of MB type of base macroblock on the base_mode_flag, we have analyzed the**probability**distribution depending on MB type of base macroblock and the current context model using Palma test condition. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% FootballBusForemanMobile ctxIdxInc2,flagOff,baseInter ctxIdxInc2,flagOff,baseIntra ctxIdxInc1,flagOff,baseInter ctxIdxInc1,flagOff,baseIntra ctxIdxInc0,flagOff,baseInter ctxIdxInc0,flagOff,baseIntra ctxIdxInc2,flagOn,baseInter ctxIdxInc2,flagOn,baseIntra ctxIdxInc1,flagOn,baseInter ctxIdxInc1,flagOn,baseIntra ctxIdxInc0,flagOn,baseInter ctxIdxInc0,flagOn,baseIntra Fig.8**Probability**distribution according to MB type of base macroblock and context model 0% 20% 40% 60% 80% 100% FootballBusForemanMobile ctxIdxInc0,flagOff,baseIntra ctxIdxInc0,flagOn,baseIntra Fig.9**Probability**distribution when base macroblock is intra and ctxIdxInc = 0 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% FootballBusForemanMobile ctxIdxInc0,flagOff,baseInter ctxIdxInc0,flagOn,baseInter Fig.10**Probability**distribution when base macroblock is inter and ctxIdxInc = 0 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% FootballBusForemanMobile ctxIdxInc1,flagOff,baseIntra ctxIdxInc1,flagOn,baseIntra Fig.11**Probability**distribution when base macroblock is intra and ctxIdxInc = 1 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% FootballBusForemanMobile ctxIdxInc1,flagOff,baseInter ctxIdxInc1,flagOn,baseInter Fig.12**Probability**distribution when base macroblock is inter and ctxIdxInc = 1 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 1234 ctxIdxInc2,flagOff,baseIntra ctxIdxInc2,flagOn,baseIntra Fig.13**Probability**distribution when base macroblock is intra and ctxIdxInc = 2 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 1234 ctxIdxInc2,flagOff,baseInter ctxIdxInc2,flagOn,baseInter Fig.14**Probability**distribution when base macroblock is inter and ctxIdxInc = 2 As shown figures, if MB type of base macroblock is inter, the**probability**of base_mode_flag is very similar to that of JSVM 2.0.**Language:**English

**Score:**730622.85 - https://www.itu.int/wftp3/av-a...te/2005_07_Poznan/JVT-P104.doc

**Data Source:**un