[x265] [PATCH] Add: Fix for VMAF Exceptions
Dakshinya T R S
dakshinya at multicorewareinc.com
Sat Apr 17 10:48:26 UTC 2021
>From e4401e650d05cb86aadcc4d22716b5ce6b6e1e84 Mon Sep 17 00:00:00 2001
From: Dakshinya T R S <dakshinya at multicorewareinc.com>
Date: Fri, 16 Apr 2021 10:40:00 +0530
Subject: [PATCH] Fix VMAF exceptions
---
doc/reST/api.rst | 2 ++
source/CMakeLists.txt | 2 +-
source/encoder/api.cpp | 8 +++++---
source/encoder/encoder.cpp | 20 ++++++++++++++++++++
source/encoder/encoder.h | 4 ++++
source/encoder/frameencoder.cpp | 1 +
source/x265.h | 5 +++--
7 files changed, 36 insertions(+), 6 deletions(-)
diff --git a/doc/reST/api.rst b/doc/reST/api.rst
index 7444f35af..96a62c40f 100644
--- a/doc/reST/api.rst
+++ b/doc/reST/api.rst
@@ -424,6 +424,8 @@ VMAF (Video Multi-Method Assessment Fusion)
If you set the ENABLE_LIBVMAF cmake option to ON, then x265 will report
per frame
and aggregate VMAF score for the given input and dump the scores in csv
file.
The user also need to specify the :option:`--recon` in command line to get
the VMAF scores.
+
+VMAF version support >= 2.0.0
/* x265_calculate_vmafScore:
* returns VMAF score for the input video.
diff --git a/source/CMakeLists.txt b/source/CMakeLists.txt
index b4e57b592..f4a9cb793 100755
--- a/source/CMakeLists.txt
+++ b/source/CMakeLists.txt
@@ -29,7 +29,7 @@ option(NATIVE_BUILD "Target the build CPU" OFF)
option(STATIC_LINK_CRT "Statically link C runtime for release builds" OFF)
mark_as_advanced(FPROFILE_USE FPROFILE_GENERATE NATIVE_BUILD)
# X265_BUILD must be incremented each time the public API is changed
-set(X265_BUILD 200)
+set(X265_BUILD 201)
configure_file("${PROJECT_SOURCE_DIR}/x265.def.in"
"${PROJECT_BINARY_DIR}/x265.def")
configure_file("${PROJECT_SOURCE_DIR}/x265_config.h.in"
diff --git a/source/encoder/api.cpp b/source/encoder/api.cpp
index 66f26e37e..5096e52ab 100644
--- a/source/encoder/api.cpp
+++ b/source/encoder/api.cpp
@@ -633,7 +633,9 @@ void x265_vmaf_encoder_log(x265_encoder* enc, int argc,
char **argv, x265_param
if (enc)
{
Encoder *encoder = static_cast<Encoder*>(enc);
- x265_stats stats;
+ x265_stats stats;
+
+ vmafdata->format = encoder->m_vmafInputFormat;
stats.aggregateVmafScore = x265_calculate_vmafscore(param,
vmafdata);
if(vmafdata->reference_file)
fclose(vmafdata->reference_file);
@@ -1925,7 +1927,7 @@ double x265_calculate_vmafscore(x265_param *param,
x265_vmaf_data *data)
else
x265_log(NULL, X265_LOG_ERROR, "Invalid format\n");
- compute_vmaf(&score, vcd->format, data->width, data->height,
read_frame, data, vcd->model_path, vcd->log_path, vcd->log_fmt,
vcd->disable_clip, vcd->disable_avx, vcd->enable_transform,
vcd->phone_model, vcd->psnr, vcd->ssim, vcd->ms_ssim, vcd->pool,
vcd->thread, vcd->subsample, vcd->enable_conf_interval);
+ compute_vmaf(&score, data->format, data->width, data->height,
read_frame, data, vcd->model_path, vcd->log_path, vcd->log_fmt,
vcd->disable_clip, vcd->disable_avx, vcd->enable_transform,
vcd->phone_model, vcd->psnr, vcd->ssim, vcd->ms_ssim, vcd->pool,
vcd->thread, vcd->subsample, vcd->enable_conf_interval);
return score;
}
@@ -2027,7 +2029,7 @@ double
x265_calculate_vmaf_framelevelscore(x265_vmaf_framedata *vmafframedata)
read_frame = read_frame_8bit;
else
read_frame = read_frame_10bit;
- compute_vmaf(&score, vcd->format, vmafframedata->width,
vmafframedata->height, read_frame, vmafframedata, vcd->model_path,
vcd->log_path, vcd->log_fmt, vcd->disable_clip, vcd->disable_avx,
vcd->enable_transform, vcd->phone_model, vcd->psnr, vcd->ssim,
vcd->ms_ssim, vcd->pool, vcd->thread, vcd->subsample,
vcd->enable_conf_interval);
+ compute_vmaf(&score, vmafframedata->format, vmafframedata->width,
vmafframedata->height, read_frame, vmafframedata, vcd->model_path,
vcd->log_path, vcd->log_fmt, vcd->disable_clip, vcd->disable_avx,
vcd->enable_transform, vcd->phone_model, vcd->psnr, vcd->ssim,
vcd->ms_ssim, vcd->pool, vcd->thread, vcd->subsample,
vcd->enable_conf_interval);
return score;
}
diff --git a/source/encoder/encoder.cpp b/source/encoder/encoder.cpp
index c1e1cb46d..3d4742299 100644
--- a/source/encoder/encoder.cpp
+++ b/source/encoder/encoder.cpp
@@ -586,6 +586,22 @@ void Encoder::create()
fail:
m_aborted = true;
}
+#ifdef ENABLE_LIBVMAF
+ m_vmafInputFormat = X265_MALLOC(char, 11);
+ if (m_param->internalCsp == X265_CSP_I420)
+ memcpy(m_vmafInputFormat, "yuv420p", 7);
+ else if (m_param->internalCsp == X265_CSP_I422)
+ memcpy(m_vmafInputFormat, "yuv422p", 7);
+ else if (m_param->internalCsp == X265_CSP_I444)
+ memcpy(m_vmafInputFormat, "yuv444p", 7);
+ else
+ x265_log(NULL, X265_LOG_ERROR, "Invalid format\n");
+
+ if (m_param->internalBitDepth == 10)
+ strcat(m_vmafInputFormat, "10le");
+ else if (m_param->internalBitDepth == 12)
+ strcat(m_vmafInputFormat, "12le");
+#endif
}
void Encoder::stopJobs()
@@ -999,6 +1015,10 @@ void Encoder::destroy()
#ifdef SVT_HEVC
X265_FREE(m_svtAppData);
#endif
+
+#ifdef ENABLE_LIBVMAF
+ X265_FREE(m_vmafInputFormat);
+#endif
if (m_param)
{
if (m_param->csvfpt)
diff --git a/source/encoder/encoder.h b/source/encoder/encoder.h
index 2ee5bdaee..d0b74db70 100644
--- a/source/encoder/encoder.h
+++ b/source/encoder/encoder.h
@@ -296,6 +296,10 @@ public:
ThreadSafeInteger* zoneReadCount;
ThreadSafeInteger* zoneWriteCount;
+#ifdef ENABLE_LIBVMAF
+ char* m_vmafInputFormat;
+#endif
+
Encoder();
~Encoder()
{
diff --git a/source/encoder/frameencoder.cpp
b/source/encoder/frameencoder.cpp
index 1b3875a25..11221078d 100644
--- a/source/encoder/frameencoder.cpp
+++ b/source/encoder/frameencoder.cpp
@@ -2151,6 +2151,7 @@ void FrameEncoder::vmafFrameLevelScore()
vmafframedata->internalBitDepth = m_param->internalBitDepth;
vmafframedata->reference_frame = fenc;
vmafframedata->distorted_frame = recon;
+ vmafframedata->format = m_top->m_vmafInputFormat;
fenc->m_vmafScore = x265_calculate_vmaf_framelevelscore(vmafframedata);
diff --git a/source/x265.h b/source/x265.h
index 324f3163f..a37741f53 100644
--- a/source/x265.h
+++ b/source/x265.h
@@ -712,6 +712,7 @@ typedef struct x265_vmaf_data
int internalBitDepth;
FILE *reference_file; /* FILE pointer for input file */
FILE *distorted_file; /* FILE pointer for recon file generated*/
+ char *format;
}x265_vmaf_data;
/* data to calculate frame level VMAF score */
@@ -723,12 +724,12 @@ typedef struct x265_vmaf_framedata
int internalBitDepth;
void *reference_frame; /* points to fenc of particular frame */
void *distorted_frame; /* points to recon of particular frame */
+ char *format;
}x265_vmaf_framedata;
/* common data needed to calculate both frame level and video level VMAF
scores */
typedef struct x265_vmaf_commondata
{
- char *format;
char *model_path;
char *log_path;
char *log_fmt;
@@ -745,7 +746,7 @@ typedef struct x265_vmaf_commondata
int enable_conf_interval;
}x265_vmaf_commondata;
-static const x265_vmaf_commondata vcd[] = { { NULL, (char
*)"/usr/local/share/model/vmaf_v0.6.1.pkl", NULL, NULL, 0, 0, 0, 0, 0, 0,
0, NULL, 0, 1, 0 } };
+static const x265_vmaf_commondata vcd[] = { { (char
*)"/usr/local/share/model/vmaf_v0.6.1.json", NULL, NULL, 0, 0, 0, 0, 0, 0,
0, NULL, 0, 1, 0 } };
/* x265 input parameters
*
--
2.17.1
Thanks & Regards,
Dakshinya T R S,
Intern at Multicoreware,
Bannari Amman Institute of Technology,
Sathyamangalam.
______________________________________
x265-devel mailing list
x265-devel at videolan.org
https://mailman.videolan.org/listinfo/x265-devel
On Thu, Apr 15, 2021 at 6:08 PM Dakshinya T R S <
dakshinya at multicorewareinc.com> wrote:
> Sure. I have noted it and I am working on the changes.
>
> Thanks & Regards,
> Dakshinya T R S,
> Intern at Multicoreware,
> Bannari Amman Institute of Technology,
> Sathyamangalam.
>
>
>
>
> On Mon, Apr 12, 2021 at 4:44 PM Pooja Venkatesan <
> pooja at multicorewareinc.com> wrote:
>
>> FYI
>> Regards,
>> *Pooja Venkatesan*,
>> Video Codec Engineer,
>> Media & AI analytics BU
>>
>>
>>
>> ---------- Forwarded message ---------
>> From: Aruna Matheswaran <aruna at multicorewareinc.com>
>> Date: Mon, Apr 12, 2021 at 3:54 PM
>> Subject: Re: [x265] [PATCH] Add: Fix for VMAF Exceptions
>> To: Development for x265 <x265-devel at videolan.org>
>>
>>
>>
>>
>> On Fri, Apr 9, 2021 at 12:35 PM Dakshinya T R S <
>> dakshinya at multicorewareinc.com> wrote:
>>
>>> From bf71142a6b22d33b2b2b8ebc8d80a019dbdf0a1d Mon Sep 17 00:00:00 2001
>>> From: Dakshinya T R S <dakshinya at multicorewareinc.com>
>>> Date: Wed, 7 Apr 2021 23:32:15 -0700
>>> Subject: [PATCH] Fix VMAF exceptions
>>>
>>> ---
>>> source/encoder/api.cpp | 38 ++++++++++++++++++++++++++++++---
>>> source/encoder/frameencoder.cpp | 1 +
>>> source/x265.h | 10 +++++----
>>> 3 files changed, 42 insertions(+), 7 deletions(-)
>>>
>>> diff --git a/source/encoder/api.cpp b/source/encoder/api.cpp
>>> index a986355e0..9374bbc38 100644
>>> --- a/source/encoder/api.cpp
>>> +++ b/source/encoder/api.cpp
>>> @@ -1903,22 +1903,34 @@ double x265_calculate_vmafscore(x265_param
>>> *param, x265_vmaf_data *data)
>>> data->width = param->sourceWidth;
>>> data->height = param->sourceHeight;
>>> data->internalBitDepth = param->internalBitDepth;
>>> + data->format = (char *)malloc(11 * sizeof(char));
>>> [AM] Repeated malloc shall be avoided if the format is captured in
>>> encoder handle or in m_cliopt.vmafData after input file info parsing.
>>> if (param->internalCsp == X265_CSP_I420)
>>> {
>>> if ((param->sourceWidth * param->sourceHeight) % 2 != 0)
>>> x265_log(NULL, X265_LOG_ERROR, "Invalid file size\n");
>>> data->offset = param->sourceWidth * param->sourceHeight / 2;
>>> + memcpy(data->format, "yuv420p", 7);
>>> }
>>> else if (param->internalCsp == X265_CSP_I422)
>>> + {
>>> data->offset = param->sourceWidth * param->sourceHeight;
>>> + memcpy(data->format, "yuv422p", 7);
>>> + }
>>> else if (param->internalCsp == X265_CSP_I444)
>>> + {
>>> data->offset = param->sourceWidth * param->sourceHeight * 2;
>>> + memcpy(data->format, "yuv444p", 7);
>>> + }
>>> else
>>> x265_log(NULL, X265_LOG_ERROR, "Invalid format\n");
>>> + if (param->internalBitDepth == 10)
>>> + strcat(data->format, "10le");
>>> + else if (param->internalBitDepth == 12)
>>> + strcat(data->format, "12le");
>>>
>>> - compute_vmaf(&score, vcd->format, data->width, data->height,
>>> read_frame, data, vcd->model_path, vcd->log_path, vcd->log_fmt,
>>> vcd->disable_clip, vcd->disable_avx, vcd->enable_transform,
>>> vcd->phone_model, vcd->psnr, vcd->ssim, vcd->ms_ssim, vcd->pool,
>>> vcd->thread, vcd->subsample, vcd->enable_conf_interval);
>>> -
>>> + compute_vmaf(&score, data->format, data->width, data->height,
>>> read_frame, data, vcd->model_path, vcd->log_path, vcd->log_fmt,
>>> vcd->disable_clip, vcd->disable_avx, vcd->enable_transform,
>>> vcd->phone_model, vcd->psnr, vcd->ssim, vcd->ms_ssim, vcd->pool,
>>> vcd->thread, vcd->subsample, vcd->enable_conf_interval);
>>> + free(data->format);
>>> return score;
>>> }
>>>
>>> @@ -2019,8 +2031,28 @@ double
>>> x265_calculate_vmaf_framelevelscore(x265_vmaf_framedata *vmafframedata)
>>> read_frame = read_frame_8bit;
>>> else
>>> read_frame = read_frame_10bit;
>>> - compute_vmaf(&score, vcd->format, vmafframedata->width,
>>> vmafframedata->height, read_frame, vmafframedata, vcd->model_path,
>>> vcd->log_path, vcd->log_fmt, vcd->disable_clip, vcd->disable_avx,
>>> vcd->enable_transform, vcd->phone_model, vcd->psnr, vcd->ssim,
>>> vcd->ms_ssim, vcd->pool, vcd->thread, vcd->subsample,
>>> vcd->enable_conf_interval);
>>> + vmafframedata->format = (char *)malloc(11 * sizeof(char));
>>>
>>> + if (vmafframedata->internalCsp == X265_CSP_I420)
>>> + {
>>> + memcpy(vmafframedata->format, "yuv420p", 7);
>>> + }
>>> + else if (vmafframedata->internalCsp == X265_CSP_I422)
>>> + {
>>> + memcpy(vmafframedata->format, "yuv422p", 7);
>>> + }
>>> + else if (vmafframedata->internalCsp == X265_CSP_I444)
>>> + {
>>> + memcpy(vmafframedata->format, "yuv444p", 7);
>>> + }
>>> + else
>>> + x265_log(NULL, X265_LOG_ERROR, "Invalid format\n");
>>> + if (vmafframedata->internalBitDepth == 10)
>>> + strcat(vmafframedata->format, "10le");
>>> + else if (vmafframedata->internalBitDepth == 12)
>>> + strcat(vmafframedata->format, "12le");
>>>
>> [AM] Same comment as above.
>>
>>> + compute_vmaf(&score, vmafframedata->format, vmafframedata->width,
>>> vmafframedata->height, read_frame, vmafframedata, vcd->model_path,
>>> vcd->log_path, vcd->log_fmt, vcd->disable_clip, vcd->disable_avx,
>>> vcd->enable_transform, vcd->phone_model, vcd->psnr, vcd->ssim,
>>> vcd->ms_ssim, vcd->pool, vcd->thread, vcd->subsample,
>>> vcd->enable_conf_interval);
>>> + free(vmafframedata->format);
>>> return score;
>>> }
>>> #endif
>>> diff --git a/source/encoder/frameencoder.cpp
>>> b/source/encoder/frameencoder.cpp
>>> index 2086a15a5..05c1d93a5 100644
>>> --- a/source/encoder/frameencoder.cpp
>>> +++ b/source/encoder/frameencoder.cpp
>>> @@ -2143,6 +2143,7 @@ void FrameEncoder::vmafFrameLevelScore()
>>> vmafframedata->internalBitDepth = m_param->internalBitDepth;
>>> vmafframedata->reference_frame = fenc;
>>> vmafframedata->distorted_frame = recon;
>>> + vmafframedata->internalCsp = m_param->internalCsp;
>>>
>>> fenc->m_vmafScore =
>>> x265_calculate_vmaf_framelevelscore(vmafframedata);
>>>
>>> diff --git a/source/x265.h b/source/x265.h
>>> index f44040ba7..0e572e468 100644
>>> --- a/source/x265.h
>>> +++ b/source/x265.h
>>> @@ -709,6 +709,7 @@ typedef struct x265_vmaf_data
>>> int internalBitDepth;
>>> FILE *reference_file; /* FILE pointer for input file */
>>> FILE *distorted_file; /* FILE pointer for recon file generated*/
>>> + char *format;
>>> }x265_vmaf_data;
>>> [AM] Update X265_BUILD.
>>> /* data to calculate frame level VMAF score */
>>> @@ -716,16 +717,17 @@ typedef struct x265_vmaf_framedata
>>> {
>>> int width;
>>> int height;
>>> - int frame_set;
>>> - int internalBitDepth;
>>> + int frame_set;
>>> + int internalBitDepth;
>>> void *reference_frame; /* points to fenc of particular frame */
>>> void *distorted_frame; /* points to recon of particular frame */
>>> + char *format;
>>> + int internalCsp;
>>> }x265_vmaf_framedata;
>>>
>>> /* common data needed to calculate both frame level and video level
>>> VMAF scores */
>>> typedef struct x265_vmaf_commondata
>>> {
>>> - char *format;
>>> char *model_path;
>>> char *log_path;
>>> char *log_fmt;
>>> @@ -742,7 +744,7 @@ typedef struct x265_vmaf_commondata
>>> int enable_conf_interval;
>>> }x265_vmaf_commondata;
>>>
>>> -static const x265_vmaf_commondata vcd[] = { { NULL, (char
>>> *)"/usr/local/share/model/vmaf_v0.6.1.pkl", NULL, NULL, 0, 0, 0, 0, 0, 0,
>>> 0, NULL, 0, 1, 0 } };
>>> +static const x265_vmaf_commondata vcd[] = { { (char
>>> *)"/usr/local/share/model/vmaf_v0.6.1.json", NULL, NULL, 0, 0, 0, 0, 0, 0,
>>> 0, NULL, 0, 1, 0 } };
>>>
>>> /* x265 input parameters
>>>
>> [AM] Please mention the VMAF version with the integration works fine in
>> VMAF documentation.
>>
>>> *
>>> --
>>> 2.17.1
>>>
>>>
>>> Thanks & Regards,
>>> Dakshinya T R S,
>>> Intern at Multicoreware,
>>> Bannari Amman Institute of Technology,
>>> Sathyamangalam.
>>> ______________________________________
>>>
>>> x265-devel mailing list
>>> x265-devel at videolan.org
>>> https://mailman.videolan.org/listinfo/x265-devel
>>> _______________________________________________
>>> x265-devel mailing list
>>> x265-devel at videolan.org
>>> https://mailman.videolan.org/listinfo/x265-devel
>>>
>>
>>
>> --
>> Regards,
>> *Aruna Matheswaran,*
>> Video Codec Engineer,
>> Media & AI analytics BU,
>>
>>
>>
>> _______________________________________________
>> x265-devel mailing list
>> x265-devel at videolan.org
>> https://mailman.videolan.org/listinfo/x265-devel
>>
>
--
*Dakshinya T R S,*
*Intern at Multicoreware,*
*Bannari Amman Institute of Technology,*
*Sathyamangalam.*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.videolan.org/pipermail/x265-devel/attachments/20210417/b700851a/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Fix-VMAF-exceptions.patch
Type: application/octet-stream
Size: 7522 bytes
Desc: not available
URL: <http://mailman.videolan.org/pipermail/x265-devel/attachments/20210417/b700851a/attachment-0001.obj>
More information about the x265-devel
mailing list