generic-library/vpx
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
18,311 Commits 55 Branches 16 Tags
Commit Graph

748 Commits

Author SHA1 Message Date
Jingning Han
b5a3082190 Merge "Synchronize tx_size counts in the decoder" into nextgenv2 2016-10-26 21:46:18 +00:00
Ryan Lei
6f8c1a78da Fix two bugs in parallel_deblocking experiment 
This commit fixes two major bugs in parallel deblocking experiment, the
first one is missing initialization of lfm->lfl_uv array for horizontal
filtering. The second one is inconsistent order of vertical/horizontal
filtering of superblocks within a frame between encoder and decoder.

BUG=https://bugs.chromium.org/p/aomedia/issues/detail?id=45#c2
BUG=https://bugs.chromium.org/p/aomedia/issues/detail?id=53#c1

Change-Id: I2df7eb313d49203fb70efe2bdf957b9d7e0bf678
 2016-10-26 13:42:31 -07:00
Sarah Parker
7ba8dc1688 Fix ubsan left shift warnings in warped motion library 
Change-Id: I14f609664411577706dbe4c099d90f0cfe2f7bb3
 2016-10-26 12:58:36 -07:00
Sarah Parker
b3dab4983b Cosmetic fixes in global motion experiment 
These are in response to post-commit suggestions made on
If429c93bb90b66fdff0edc07ecd9fc078077d303.

Change-Id: Id29afa158471bd6259bd07ac00812a50bfd0a709
 2016-10-26 11:45:50 -07:00
Urvang Joshi
839b07feec Merge changes I56cddcb4,I40c5a652 into nextgenv2 
* changes:
  Mark bogus palette color probabilities as zero
  get_palette_color_context: code cleanup
 2016-10-26 18:28:56 +00:00
Jingning Han
906be078a5 Synchronize tx_size counts in the decoder 
Make both encoder and decoder use depth index for frame count.

Change-Id: I96dddffc0a83ad5e4e2847b15391e01ba01ee502
 2016-10-26 11:04:58 -07:00
Angie Chiang
180566d854 Merge "av1/convolve.[hc],av1_convolve_test: add missing copyright" into nextgenv2 2016-10-26 17:51:28 +00:00
Angie Chiang
8e26f768c1 Merge "Use has_subpel_mv_component in av1_is_interp_needed" into nextgenv2 2016-10-26 17:50:54 +00:00
Angie Chiang
65eb2cf78a Sync definition of av1_get_switchable_rate in rd.c/h 
Change-Id: I720934e02a15fd6184bdda6c1b8a23d5b02a5284
 2016-10-26 10:48:47 -07:00
Sarah Parker
70c0df29da Merge "Revise precision clamping in GM param refinement" into nextgenv2 2016-10-26 17:33:47 +00:00
Jingning Han
b0a7130656 Convert tx_size to relative depth to fetch tx_size_cost 
Use the relative transform partition depth as index to fetch the
tx_size_cost value.

Change-Id: I7d5119817baa96f23c32828065ff3175bb9f75cf
 2016-10-26 17:12:41 +00:00
Jingning Han
8e022edd59 Allow backward probability model update from tx_size=0 
Replace enum items with range definitions.

Change-Id: Iba2b7cac657db5fb6177cb5c9e6f40ec0125d926
 2016-10-26 17:12:20 +00:00
Jingning Han
e5596d3168 Merge "Add depth to tx_size mapper to bit-stream coding" into nextgenv2 2016-10-26 17:11:56 +00:00
Angie Chiang
c352e79ee6 Merge "Simplify interpolation filter search in handle_inter_mode()" into nextgenv2 2016-10-26 16:51:58 +00:00
Janne Salonen
e8a3dbc0ff fix filtering of uv int4x4 for odd rows 
Change-Id: I61f91855430e11da45d4e91ec6d3a8976c461cb7
 2016-10-26 09:26:28 -07:00
Ryan Lei
15149484ec Add parallel-deblocking experiment 
This commit is a manual cherry-pick from aom/master:
42ff3881ace1564aac9debae86ef37a8deb8d381

Change-Id: I4a3cdb939b7b96a3aa27f6a00da7a0e73222f3f3
 2016-10-26 09:20:47 -07:00
Yunqing Wang
e61ec7bc19 Merge "Change 2 motion search counts to be tile data" into nextgenv2 2016-10-26 16:17:42 +00:00
Yaowu Xu
5a1fedfdda av1/common/filter.h: apply clang-format 
Change-Id: I37f0d1fbcc6f262ae287290e2e6f5648ad0113c8
 2016-10-26 09:14:01 -07:00
Jingning Han
4e1737af64 Add depth to tx_size mapper to bit-stream coding 
It serves as a helper function to integrate various transform coding
options.

Change-Id: I64e7d0c88ea10137fa1ff1072d865eb0054c2a25
 2016-10-26 15:45:19 +00:00
Sarah Parker
f41a06b231 Revise precision clamping in GM param refinement 
This ensures that the parameter refinement never
results in a motion parameter value that exceeds the number
of alloted bits in the bitstream. It accounts for all of
the necessary precision shifts required to make global motion compatible
with the warped motion library. It also accounts for the
zero-centering that is applied to global motion parameters that are
naturally centered around one.
Change-Id: If429c93bb90b66fdff0edc07ecd9fc078077d303
 2016-10-25 21:11:39 -07:00
Jingning Han
c83ef8b946 Merge "Refactor transform size coding" into nextgenv2 2016-10-26 01:12:04 +00:00
Angie Chiang
a2b56d3e05 Refactor: Localize tmp_rd in handle_inter_mode() 
Change-Id: I01cb5cd544c849be160a9441d141c01a3424d32b
 2016-10-25 17:34:59 -07:00
Angie Chiang
b135debcb6 Use has_subpel_mv_component in av1_is_interp_needed 
Change-Id: I8980df4512de605aaa6a67c1f05e544f69a12e96
 2016-10-25 17:10:19 -07:00
Angie Chiang
75c2209341 Simplify interpolation filter search in handle_inter_mode() 
BDRate
ext_interp  lowres -0.001%
dual_filter lowres  0.001%

Change-Id: Ic24165d554c300eaa0188ee8cb88d320b74125aa
 2016-10-25 17:10:08 -07:00
Angie Chiang
6421191247 av1/convolve.[hc],av1_convolve_test: add missing copyright 
Change-Id: Ie84bdf90c31b12977d32baacfc8086c1fdd96e65
 2016-10-25 16:43:43 -07:00
Jingning Han
aae72a69c3 Refactor transform size coding 
Introduce the transform block partition depth macro definition.

Change-Id: I218dc77a77c8e967da4d270d4ec0d7691b712a5f
 2016-10-25 15:42:30 -07:00
Jingning Han
b2d6a59ad5 Merge "Refactor tx_size use case in block encoding stage" into nextgenv2 2016-10-25 22:29:21 +00:00
Jingning Han
2eded9a3ff Merge "Refactor tokenize_vartx to use aligned transform block size fetch" into nextgenv2 2016-10-25 22:29:03 +00:00
Yunqing Wang
8c1e57c278 Change 2 motion search counts to be tile data 
Imported changes from VP9:
https://chromium-review.googlesource.com/#/c/402551/
https://chromium-review.googlesource.com/#/c/403128/

Change-Id: I8570c867190a6fa641926431ce97f7d9d7da3528
 2016-10-25 15:25:37 -07:00
Jingning Han
a1730659ec Merge "Use table fetch for block width in block_rd_txfm" into nextgenv2 2016-10-25 22:18:44 +00:00
James Zern
8aa4cbf5d5 Merge "update_state: quiet const warning w/global-motion" into nextgenv2 2016-10-25 22:15:39 +00:00
Yaowu Xu
c2ac0a1d4c Merge "7-bit interpolation filters" into nextgenv2 2016-10-25 19:41:07 +00:00
Yaowu Xu
dece603fdf Merge "Use constrained tokenset with --enable-daala_ec." into nextgenv2 2016-10-25 19:40:51 +00:00
Jingning Han
e71ad1d4a2 Merge "Refactor dist_block() function" into nextgenv2 2016-10-25 19:39:22 +00:00
Angie Chiang
7c7e555ca0 Merge changes I6faedb29,Ic6586114 into nextgenv2 
* changes:
  Remove speed feature of ext_interp experiment
  Refactor: handle_inter()
 2016-10-25 19:36:49 +00:00
Jingning Han
de953b9d05 Refactor tx_size use case in block encoding stage 
Change-Id: I56110d1fc94b335668e6b991442e9083bbaea8ee
 2016-10-25 12:36:09 -07:00
Jingning Han
a893936335 Refactor tokenize_vartx to use aligned transform block size fetch 
This prepares for the integration of rectangular transform size
into recursive transform block partition.

Change-Id: I164eb43d10afa9bb2f4722de7a48faa770ba4ced
 2016-10-25 12:16:21 -07:00
Jingning Han
99e7a8d837 Merge "Refactor tx_size use cases in blockd.c" into nextgenv2 2016-10-25 19:03:29 +00:00
Jingning Han
c598cf853f Use table fetch for block width in block_rd_txfm 
Make direct use of block_size_wide to fetch data for stride.

Change-Id: I0d8491e58cf00ea73c764d218cb56408b64d9ee7
 2016-10-25 10:47:46 -07:00
Jingning Han
b9c572706d Refactor dist_block() function 
Support automatic scale for mapping between transform block size
and pixel block size.

Change-Id: I141b0477a85c0dcc5f99b4e5d880cfccfae6d316
 2016-10-25 10:22:17 -07:00
Arild Fuldseth
7acfabbc40 7-bit interpolation filters 
Purpose:
-Reduce dynamic range of interpolation filter coefficents from 8
bits to 7 bits.
-Inner product for 8-bit input data can be stored in a 16-bit signed
integer.

Impact on compression efficiency:
-Marginal improvement, typically less than 0.5% BDR.

Change-Id: I58d1408307ae7d2a6f9de8965c5877b258703199
 2016-10-25 10:18:55 -07:00
Yaowu Xu
1f112841d2 Merge "Refactor extrabits packing" into nextgenv2 2016-10-25 17:14:57 +00:00
Yaowu Xu
d8dc1fc522 Merge "Linearize extrabits writing." into nextgenv2 2016-10-25 17:14:44 +00:00
Nathan E. Egge
46e8490498 Use constrained tokenset with --enable-daala_ec. 
Change-Id: Ia09edf92bf9f7ecacc65c232ac6e656cde236634
 2016-10-25 10:13:22 -07:00
Jingning Han
95cff5c979 Refactor tx_size use cases in blockd.c 
Use table to replace the arithmetic computation for mapping between
transform block and pixel number. Support automatic scale of block
size and transform block size.

Change-Id: I84766850172265d4295f418383dbc5e6e5838ec8
 2016-10-25 09:50:07 -07:00
Debargha Mukherjee
7f9eb87082 Merge "Fix compile error with --enable-accounting" into nextgenv2 2016-10-25 16:25:42 +00:00
Angie Chiang
d35e12b184 Merge "Refactor: Add macro LOG_SWITCHABLE_FILTERS" into nextgenv2 2016-10-25 16:24:23 +00:00
Angie Chiang
d0aa90ed79 Remove speed feature of ext_interp experiment 
This is to facilitate the refactor process

Change-Id: I6faedb29129b47abefe20821dc3f32a43db149d8
 2016-10-25 09:22:35 -07:00
Angie Chiang
6305abe114 Refactor: Add macro LOG_SWITCHABLE_FILTERS 
Change-Id: I7593ff2f8949d8bc26ca1c8577faaefb09640b59
 2016-10-25 09:22:35 -07:00
Angie Chiang
1b131f1c64 Refactor: handle_inter() 
Make the parenthesis symmetric
Replace interpolation filter mode number by macro

Change-Id: Ic6586114c4cebe920b950e1b3adc8ebc764d4713
 2016-10-25 09:22:35 -07:00