New motion vector entropy coding

Adds a new experiment with redesigned/refactored motion vector entropy
coding. The patch also takes a first step towards separating the
integer and fractional pel components of a MV. However the fractional
pel encoding still depends on the integer pel part and so they are
not fully independent. Further experiments are in progress to see
how much they can be decoupled without affecting performance.
All components including entropy coding/decoding, costing for MV
search, forward updates and backward updates to probability tables,
have been implemented.

Results so far:
derf: +0.19%
std-hd: +0.28%
yt: +0.80%
hd: +1.15%

Patch: Simplifies the fractional pel models:
derf: +0.284%
std-hd: +0.289%
yt: +0.849%
hd: +1.254%

Patch: Some changes in the models, rebased.
derf: +0.330%
std-hd: +0.306%
yt: +0.816%
hd: +1.225%

Change-Id: I646b3c48f3587f4cc909639b78c3798da6402678

1 files changed, 13 insertions, 3 deletions

diff --git a/vp8/decoder/decodframe.c b/vp8/decoder/decodframe.c
index d93268396..5fb510b64 100644
--- a/vp8/decoder/decodframe.c
+++ b/vp8/decoder/decodframe.c
@@ -923,9 +923,7 @@ static void init_frame(VP8D_COMP *pbi) {
 
   if (pc->frame_type == KEY_FRAME) {
     /* Various keyframe initializations */
-    vpx_memcpy(pc->fc.mvc, vp8_default_mv_context, sizeof(vp8_default_mv_context));
-    vpx_memcpy(pc->fc.mvc_hp, vp8_default_mv_context_hp,
-               sizeof(vp8_default_mv_context_hp));
+    vp8_init_mv_probs(pc);
 
     vp8_init_mbmode_probs(pc);
 
@@ -1464,8 +1462,12 @@ int vp8_decode_frame(VP8D_COMP *pbi) {
   vp8_copy(pbi->common.fc.pre_i8x8_mode_prob, pbi->common.fc.i8x8_mode_prob);
   vp8_copy(pbi->common.fc.pre_sub_mv_ref_prob, pbi->common.fc.sub_mv_ref_prob);
   vp8_copy(pbi->common.fc.pre_mbsplit_prob, pbi->common.fc.mbsplit_prob);
+#if CONFIG_NEWMVENTROPY
+  pbi->common.fc.pre_nmvc = pbi->common.fc.nmvc;
+#else
   vp8_copy(pbi->common.fc.pre_mvc, pbi->common.fc.mvc);
   vp8_copy(pbi->common.fc.pre_mvc_hp, pbi->common.fc.mvc_hp);
+#endif
   vp8_zero(pbi->common.fc.coef_counts);
   vp8_zero(pbi->common.fc.coef_counts_8x8);
 #if CONFIG_TX16X16 || CONFIG_HYBRIDTRANSFORM16X16
@@ -1477,8 +1479,12 @@ int vp8_decode_frame(VP8D_COMP *pbi) {
   vp8_zero(pbi->common.fc.i8x8_mode_counts);
   vp8_zero(pbi->common.fc.sub_mv_ref_counts);
   vp8_zero(pbi->common.fc.mbsplit_counts);
+#if CONFIG_NEWMVENTROPY
+  vp8_zero(pbi->common.fc.NMVcount);
+#else
   vp8_zero(pbi->common.fc.MVcount);
   vp8_zero(pbi->common.fc.MVcount_hp);
+#endif
   vp8_zero(pbi->common.fc.mv_ref_ct);
   vp8_zero(pbi->common.fc.mv_ref_ct_a);
 #if COEFUPDATETYPE == 2
@@ -1544,7 +1550,11 @@ int vp8_decode_frame(VP8D_COMP *pbi) {
   vp8_adapt_coef_probs(pc);
   if (pc->frame_type != KEY_FRAME) {
     vp8_adapt_mode_probs(pc);
+#if CONFIG_NEWMVENTROPY
+    vp8_adapt_nmv_probs(pc, xd->allow_high_precision_mv);
+#else
     vp8_adapt_mv_probs(pc);
+#endif
     vp8_update_mode_context(&pbi->common);
   }