class ParallelFeatureExtractorTestNew(unittest.TestCase):
def tearDown(self):
if hasattr(self, 'fextractor'):
self.fextractor.remove_results()
pass
def test_run_vmaf_fextractor_with_resampling(self):
ref_path = VmafConfig.test_resource_path("yuv", "src01_hrc00_576x324.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "src01_hrc01_576x324.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':576, 'height':324,
'quality_width':160, 'quality_height':90})
asset_original = Asset(dataset="test", content_id=0, asset_id=2,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':576, 'height':324,
'quality_width':160, 'quality_height':90})
self.fextractor = VmafFeatureExtractor(
[asset, asset_original], None, fifo_mode=False)
self.fextractor.run(parallelize=False)
results = self.fextractor.results
self.assertAlmostEqual(results[0]['VMAF_feature_vif_score'], 0.74165043750000004, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_motion_score'],1.4066421666666666, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_score'], 0.9808006041666667, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_ansnr_score'], 27.319241250000001, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_motion_score'], 1.4066421666666666, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_ansnr_score'], 30.682829895833333, places=4)
def test_run_vmaf_fextractor_with_cropping(self):
# crop_cmd: 288:162:144:81 - crop to 288x162 with upper-left pixel
# starting at coordinate (144, 81)
ref_path = VmafConfig.test_resource_path("yuv", "src01_hrc00_576x324.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "src01_hrc01_576x324.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':576, 'height':324,
'crop_cmd':'288:162:144:81',
'quality_width':288, 'quality_height':162,
})
asset_original = Asset(dataset="test", content_id=0, asset_id=2,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':576, 'height':324,
'crop_cmd':'288:162:144:81',
'quality_width':288, 'quality_height':162,
})
self.fextractor = VmafFeatureExtractor(
[asset, asset_original], None, fifo_mode=False)
self.fextractor.run(parallelize=False)
results = self.fextractor.results
self.assertAlmostEqual(results[0]['VMAF_feature_vif_score'], 0.45365762500000012, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_motion_score'], 2.8779373333333331, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm2_score'],0.93894987889874548, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_ansnr_score'], 23.942050354166668, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_motion_score'], 2.8779373333333331, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_ansnr_score'], 31.71648420833333, places=4)
def test_run_vmaf_fextractor_with_padding(self):
# pad_cmd: iw+100:ih+100:50:50 - pad to (iw+100)x(ih+100), where iw is
# input width, ih is input height, and starting point is (-50, -50)
ref_path = VmafConfig.test_resource_path("yuv", "src01_hrc00_576x324.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "src01_hrc01_576x324.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':576, 'height':324,
'pad_cmd': 'iw+100:ih+100:50:50',
'quality_width':676, 'quality_height':424,
})
asset_original = Asset(dataset="test", content_id=0, asset_id=2,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':576, 'height':324,
'pad_cmd': 'iw+100:ih+100:50:50',
'quality_width':676, 'quality_height':424,
})
self.fextractor = VmafFeatureExtractor(
[asset, asset_original], None, fifo_mode=True)
self.fextractor.run(parallelize=False)
results = self.fextractor.results
self.assertAlmostEqual(results[0]['VMAF_feature_vif_score'], 0.51023564583333325, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_motion_score'], 2.6397702083333332, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm2_score'],0.94112949409549829, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_ansnr_score'], 26.893242291666667, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_motion_score'], 2.6397702083333332, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_ansnr_score'], 34.306043416666668, places=4)
def test_run_vmaf_fextractor_with_cropping_and_padding_to_original_wh(self):
# crop_cmd: 288:162:144:81 - crop to the center 288x162 image
# pad_cmd: iw+288:ih+162:144:81 - pad back to the original size
ref_path = VmafConfig.test_resource_path("yuv", "src01_hrc00_576x324.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "src01_hrc01_576x324.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':576, 'height':324,
'crop_cmd':'288:162:144:81',
'pad_cmd': 'iw+288:ih+162:144:81',
'quality_width':576, 'quality_height':324,
})
asset_original = Asset(dataset="test", content_id=0, asset_id=2,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':576, 'height':324,
'crop_cmd':'288:162:144:81',
'pad_cmd': 'iw+288:ih+162:144:81',
'quality_width':576, 'quality_height':324,
})
self.fextractor = VmafFeatureExtractor(
[asset, asset_original], None, fifo_mode=True)
self.fextractor.run(parallelize=True)
results = self.fextractor.results
self.assertAlmostEqual(results[0]['VMAF_feature_vif_score'], 0.64106379166666672, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_motion_score'], 0.7203213958333331, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm2_score'],0.94700196017089999, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_ansnr_score'], 32.78451041666667, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_motion_score'], 0.7203213958333331, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_ansnr_score'], 40.280504208333333, places=4)
class FeatureExtractorTest(unittest.TestCase):
def tearDown(self):
if hasattr(self, 'fextractor'):
self.fextractor.remove_results()
pass
def test_executor_id(self):
asset = Asset(dataset="test", content_id=0, asset_id=1,
ref_path="dir/refvideo.yuv", dis_path="dir/disvideo.yuv",
asset_dict={'width': 720, 'height': 480})
fextractor = VmafFeatureExtractor([asset], None)
self.assertEquals(fextractor.executor_id, "VMAF_feature_V0.2.4c")
def test_get_log_file_path(self):
import hashlib
asset = Asset(dataset="test", content_id=0, asset_id=1,
ref_path="dir/refvideo.yuv", dis_path="dir/disvideo.yuv",
asset_dict={'width':720, 'height':480,},
workdir_root="my_workdir_root")
fextractor = VmafFeatureExtractor([asset], None)
log_file_path = fextractor._get_log_file_path(asset)
h = hashlib.sha1("test_0_1_refvideo_720x480_vs_disvideo_720x480_q_720x480").hexdigest()
self.assertTrue(re.match(r"^my_workdir_root/[a-zA-Z0-9-]+/VMAF_feature_V0.2.4c_{}$".format(h), log_file_path))
def test_run_vmaf_fextractor(self):
print 'test on running VMAF feature extractor...'
ref_path = VmafConfig.test_resource_path("yuv", "src01_hrc00_576x324.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "src01_hrc01_576x324.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=0,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':576, 'height':324})
asset_original = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':576, 'height':324})
self.fextractor = VmafFeatureExtractor(
[asset, asset_original],
None, fifo_mode=True,
result_store=None
)
self.fextractor.run()
results = self.fextractor.results
self.assertAlmostEqual(results[0]['VMAF_feature_vif_score'], 0.4460930625, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_motion_score'], 4.04982535417, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_motion2_score'], 3.8953518541666665, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_score'],0.93458780728708746, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm2_score'],0.93458780728708746, places=4) # at version 0.2.4b (ioannis adm fix), adm and adm2 are now identical
self.assertAlmostEqual(results[0]['VMAF_feature_ansnr_score'], 23.5095715208, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_num_score'], 712650.023478, places=0)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_den_score'], 1597314.95249, places=0)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_num_score'], 371.83541406249998, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_den_score'], 397.83378972916671, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_anpsnr_score'], 34.164776875, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_scale0_score'], 0.363420489439, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_scale1_score'], 0.766647542135, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_scale2_score'], 0.862854666902, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_scale3_score'], 0.915971778036, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale0_score'], 0.90791933424090698, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale1_score'], 0.89395660507453423, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale2_score'], 0.93010045185439161, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale3_score'], 0.96503534602850938, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif2_score'], 0.72722361912801026, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm3_score'], 0.92425293429958566, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_motion_score'], 4.04982535417, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_motion2_score'], 3.8953518541666665, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_ansnr_score'], 31.2714392708, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_num_score'], 1597314.86733, places=0)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_den_score'], 1597314.95249, places=0)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_num_score'], 397.83378972916671, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_den_score'], 397.83378972916671, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_anpsnr_score'], 41.9266444375, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_scale0_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_scale1_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_scale2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_scale3_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale0_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale1_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale3_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm3_score'], 1.0, places=4)
def test_run_vmaf_fextractor_with_result_store(self):
print 'test on running VMAF feature extractor with result store...'
ref_path = VmafConfig.test_resource_path("yuv", "src01_hrc00_576x324.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "src01_hrc01_576x324.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=0,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':576, 'height':324})
asset_original = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':576, 'height':324})
result_store = FileSystemResultStore(logger=None)
self.fextractor = VmafFeatureExtractor(
[asset, asset_original],
None, fifo_mode=True,
result_store=result_store
)
print ' running for the first time with fresh calculation...'
self.fextractor.run()
result0, result1 = self.fextractor.results
self.assertTrue(os.path.exists(result_store._get_result_file_path(result0)))
self.assertTrue(os.path.exists(result_store._get_result_file_path(result1)))
print ' running for the second time with stored results...'
self.fextractor.run()
results = self.fextractor.results
self.assertAlmostEqual(results[0]['VMAF_feature_vif_score'], 0.4460930625, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_motion_score'], 4.04982535417, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_motion2_score'], 3.8953518541666665, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm2_score'],0.93458780728708746, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_ansnr_score'], 23.5095715208, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_num_score'], 712650.023478, places=0)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_den_score'], 1597314.95249, places=0)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_num_score'], 371.83541406249998, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_den_score'], 397.83378972916671, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_anpsnr_score'], 34.164776875, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_scale0_score'], 0.363420489439, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_scale1_score'], 0.766647542135, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_scale2_score'], 0.862854666902, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_scale3_score'], 0.915971778036, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale0_score'], 0.90791933424090698, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale1_score'], 0.89395660507453423, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale2_score'], 0.93010045185439161, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale3_score'], 0.96503534602850938, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif2_score'], 0.72722361912801026, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm3_score'], 0.92425293429958566, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_motion_score'], 4.04982535417, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_motion2_score'], 3.8953518541666665, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_ansnr_score'], 31.2714392708, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_num_score'], 1597314.86733, places=0)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_den_score'], 1597314.95249, places=0)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_num_score'], 397.83378972916671, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_den_score'], 397.83378972916671, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_anpsnr_score'], 41.9266444375, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_scale0_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_scale1_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_scale2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_scale3_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale0_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale1_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale3_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm3_score'], 1.0, places=4)
def test_run_moment_fextractor(self):
print 'test on running Moment feature extractor...'
ref_path = VmafConfig.test_resource_path("yuv", "src01_hrc00_576x324.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "src01_hrc01_576x324.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=0,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':576, 'height':324})
asset_original = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':576, 'height':324})
self.fextractor = MomentFeatureExtractor(
[asset, asset_original],
None, fifo_mode=True,
result_store=None
)
self.fextractor.run()
results = self.fextractor.results
self.assertAlmostEqual(results[0]['Moment_feature_ref1st_score'], 59.788567297525134, places=4)
self.assertAlmostEqual(results[0]['Moment_feature_ref2nd_score'], 4696.668388042269, places=4)
self.assertAlmostEqual(results[0]['Moment_feature_refvar_score'], 1121.519917231203, places=4)
self.assertAlmostEqual(results[0]['Moment_feature_dis1st_score'], 61.332006624999984, places=4)
self.assertAlmostEqual(results[0]['Moment_feature_dis2nd_score'], 4798.659574041666, places=4)
self.assertAlmostEqual(results[0]['Moment_feature_disvar_score'], 1036.837184348847, places=4)
self.assertAlmostEqual(results[1]['Moment_feature_ref1st_score'], 59.788567297525134, places=4)
self.assertAlmostEqual(results[1]['Moment_feature_ref2nd_score'], 4696.668388042269, places=4)
self.assertAlmostEqual(results[1]['Moment_feature_refvar_score'], 1121.519917231203, places=4)
self.assertAlmostEqual(results[1]['Moment_feature_dis1st_score'], 59.788567297525134, places=4)
self.assertAlmostEqual(results[1]['Moment_feature_dis2nd_score'], 4696.668388042269, places=4)
self.assertAlmostEqual(results[1]['Moment_feature_disvar_score'], 1121.519917231203, places=4)
def test_run_psnr_fextractor(self):
print 'test on running PSNR feature extractor...'
ref_path = VmafConfig.test_resource_path("yuv", "src01_hrc00_576x324.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "src01_hrc01_576x324.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=0,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':576, 'height':324})
asset_original = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':576, 'height':324})
self.fextractor = PsnrFeatureExtractor(
[asset, asset_original],
None, fifo_mode=True,
result_store=None
)
self.fextractor.run()
results = self.fextractor.results
self.assertAlmostEqual(results[0]['PSNR_feature_psnr_score'], 30.755063979166664, places=4)
self.assertAlmostEqual(results[1]['PSNR_feature_psnr_score'], 60.0, places=4)
def test_run_ssim_fextractor(self):
print 'test on running SSIM feature extractor...'
ref_path = VmafConfig.test_resource_path("yuv", "src01_hrc00_576x324.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "src01_hrc01_576x324.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=0,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':576, 'height':324})
asset_original = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':576, 'height':324})
self.fextractor = SsimFeatureExtractor(
[asset, asset_original],
None, fifo_mode=True,
result_store=None
)
self.fextractor.run()
results = self.fextractor.results
self.assertAlmostEqual(results[0]['SSIM_feature_ssim_score'], 0.86322654166666657, places=4)
self.assertAlmostEqual(results[0]['SSIM_feature_ssim_l_score'], 0.9981474583333334, places=4)
self.assertAlmostEqual(results[0]['SSIM_feature_ssim_c_score'], 0.96126793750000006, places=4)
self.assertAlmostEqual(results[0]['SSIM_feature_ssim_s_score'], 0.89773633333333336, places=4)
self.assertAlmostEqual(results[1]['SSIM_feature_ssim_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['SSIM_feature_ssim_l_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['SSIM_feature_ssim_c_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['SSIM_feature_ssim_s_score'], 1.0, places=4)
def test_run_ms_ssim_fextractor(self):
print 'test on running MS-SSIM feature extractor...'
ref_path = VmafConfig.test_resource_path("yuv", "src01_hrc00_576x324.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "src01_hrc01_576x324.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=0,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':576, 'height':324})
asset_original = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':576, 'height':324})
self.fextractor = MsSsimFeatureExtractor(
[asset, asset_original],
None, fifo_mode=True,
result_store=None
)
self.fextractor.run()
results = self.fextractor.results
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_score'], 0.9632498125, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_l_scale0_score'], 0.9981474583333334, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_c_scale0_score'], 0.96126793750000006, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_s_scale0_score'], 0.89773633333333336, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_l_scale1_score'], 0.99899612500000001, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_c_scale1_score'], 0.9857694375, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_s_scale1_score'], 0.941185875, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_l_scale2_score'], 0.99923564583333324, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_c_scale2_score'], 0.997034020833, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_s_scale2_score'], 0.977992145833, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_l_scale3_score'], 0.99929210416666658, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_c_scale3_score'], 0.999588104167, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_s_scale3_score'], 0.99387125, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_l_scale4_score'], 0.99940356249999995, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_c_scale4_score'], 0.999907625, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_s_scale4_score'], 0.998222583333, places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_l_scale0_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_c_scale0_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_s_scale0_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_l_scale1_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_c_scale1_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_s_scale1_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_l_scale2_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_c_scale2_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_s_scale2_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_l_scale3_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_c_scale3_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_s_scale3_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_l_scale4_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_c_scale4_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_s_scale4_score'], 1., places=4)
def test_run_vmaf_fextractor_checkerboard(self):
print 'test on running VMAF feature extractor on checkerboard...'
ref_path = VmafConfig.test_resource_path("yuv", "checkerboard_1920_1080_10_3_0_0.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "checkerboard_1920_1080_10_3_10_0.yuv")
dis_path2 = VmafConfig.test_resource_path("yuv", "checkerboard_1920_1080_10_3_1_0.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=0,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':1920, 'height':1080})
asset_original = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':1920, 'height':1080})
asset2 = Asset(dataset="test", content_id=0, asset_id=2,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path2,
asset_dict={'width':1920, 'height':1080})
self.fextractor = VmafFeatureExtractor(
[asset, asset_original, asset2],
None, fifo_mode=True,
result_store=None
)
self.fextractor.run()
results = self.fextractor.results
self.assertAlmostEqual(results[0]['VMAF_feature_adm_score'], 0.053996333333333334, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm2_score'], 0.053996333333333334, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale0_score'], 0.23738393128710478, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale1_score'], 0.08524788663335138, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale2_score'], 0.024058909404945077, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale3_score'], 0.018034879735107798, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_motion_score'], 12.554836666666667, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_motion2_score'], 12.554836666666667, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale0_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale1_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale3_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_motion_score'], 12.554836666666667, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_motion2_score'], 12.554836666666667, places=4)
self.assertAlmostEqual(results[2]['VMAF_feature_adm_score'], 0.78533833333333336, places=4)
self.assertAlmostEqual(results[2]['VMAF_feature_adm2_score'], 0.7853384465157921, places=4)
self.assertAlmostEqual(results[2]['VMAF_feature_adm_scale0_score'], 0.72132189911792899, places=4)
self.assertAlmostEqual(results[2]['VMAF_feature_adm_scale1_score'], 0.69259738857522501, places=4)
self.assertAlmostEqual(results[2]['VMAF_feature_adm_scale2_score'], 0.80415911639244586, places=4)
self.assertAlmostEqual(results[2]['VMAF_feature_adm_scale3_score'], 0.82791889676239039, places=4)
self.assertAlmostEqual(results[2]['VMAF_feature_motion_score'], 12.554836666666667, places=4)
self.assertAlmostEqual(results[2]['VMAF_feature_motion2_score'], 12.554836666666667, places=4)
def test_run_vmaf_fextractor_flat(self):
print 'test on running VMAF feature extractor on flat pattern...'
ref_path = VmafConfig.test_resource_path("yuv", "flat_1920_1080_0.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "flat_1920_1080_10.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=0,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':1920, 'height':1080})
asset_original = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':1920, 'height':1080})
self.fextractor = VmafFeatureExtractor(
[asset, asset_original],
None, fifo_mode=True,
result_store=None
)
self.fextractor.run()
results = self.fextractor.results
self.assertAlmostEqual(results[0]['VMAF_feature_adm_score'], 1.0, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm2_score'], 1.0, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale0_score'], 1.0, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale1_score'], 1.0, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale2_score'], 1.0, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale3_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale0_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale1_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale3_score'], 1.0, places=4)
class ParallelFeatureExtractorTestNew(unittest.TestCase):
def tearDown(self):
if hasattr(self, 'fextractor'):
self.fextractor.remove_results()
pass
def test_run_vmaf_fextractor_with_gaussian_blurring(self):
ref_path = VmafConfig.test_resource_path("yuv",
"src01_hrc00_576x324.yuv")
dis_path = VmafConfig.test_resource_path("yuv",
"src01_hrc01_576x324.yuv")
asset = Asset(dataset="test",
content_id=0,
asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={
'width': 576,
'height': 324,
'crop_cmd': '288:162:144:81',
'dis_gblur_cmd': 'sigma=0.01:steps=1',
'quality_width': 288,
'quality_height': 162,
})
asset_original = Asset(dataset="test",
content_id=0,
asset_id=2,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={
'width': 576,
'height': 324,
'crop_cmd': '288:162:144:81',
'dis_gblur_cmd': 'sigma=0.01:steps=2',
'quality_width': 288,
'quality_height': 162,
})
self.fextractor = VmafFeatureExtractor([asset, asset_original],
None,
fifo_mode=True)
self.fextractor.run(parallelize=True)
results = self.fextractor.results
self.assertAlmostEqual(results[0]['VMAF_feature_vif_score'],
0.45136466666666664,
places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_motion_score'],
2.8779373333333331,
places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm2_score'],
0.9362876630569382,
places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_ansnr_score'],
24.110899416666665,
places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_score'],
0.9789283541666666,
places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_motion_score'],
2.8779373333333331,
places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm2_score'],
0.9958889086049826,
places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_ansnr_score'],
31.128021979166665,
places=4)
class FeatureExtractorTest(unittest.TestCase):
def tearDown(self):
if hasattr(self, 'fextractor'):
self.fextractor.remove_results()
pass
def test_executor_id(self):
asset = Asset(dataset="test", content_id=0, asset_id=1,
ref_path="dir/refvideo.yuv", dis_path="dir/disvideo.yuv",
asset_dict={'width': 720, 'height': 480})
fextractor = VmafFeatureExtractor([asset], None)
self.assertEquals(fextractor.executor_id, "VMAF_feature_V0.2.4b")
def test_get_log_file_path(self):
import hashlib
asset = Asset(dataset="test", content_id=0, asset_id=1,
ref_path="dir/refvideo.yuv", dis_path="dir/disvideo.yuv",
asset_dict={'width':720, 'height':480,},
workdir_root="my_workdir_root")
fextractor = VmafFeatureExtractor([asset], None)
log_file_path = fextractor._get_log_file_path(asset)
h = hashlib.sha1("test_0_1_refvideo_720x480_vs_disvideo_720x480_q_720x480").hexdigest()
self.assertTrue(re.match(r"^my_workdir_root/[a-zA-Z0-9-]+/VMAF_feature_V0.2.4b_{}$".format(h), log_file_path))
def test_run_vmaf_fextractor(self):
print 'test on running VMAF feature extractor...'
ref_path = VmafConfig.test_resource_path("yuv", "src01_hrc00_576x324.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "src01_hrc01_576x324.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=0,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':576, 'height':324})
asset_original = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':576, 'height':324})
self.fextractor = VmafFeatureExtractor(
[asset, asset_original],
None, fifo_mode=True,
result_store=None
)
self.fextractor.run()
results = self.fextractor.results
self.assertAlmostEqual(results[0]['VMAF_feature_vif_score'], 0.4460930625, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_motion_score'], 4.04982535417, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_motion2_score'], 3.8953518541666665, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_score'],0.93458780728708746, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm2_score'],0.93458780728708746, places=4) # at version 0.2.4b (ioannis adm fix), adm and adm2 are now identical
self.assertAlmostEqual(results[0]['VMAF_feature_ansnr_score'], 23.5095715208, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_num_score'], 712650.023478, places=0)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_den_score'], 1597314.95249, places=0)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_num_score'], 371.83541406249998, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_den_score'], 397.83378972916671, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_anpsnr_score'], 34.164776875, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_scale0_score'], 0.363420489439, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_scale1_score'], 0.766647542135, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_scale2_score'], 0.862854666902, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_scale3_score'], 0.915971778036, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale0_score'], 0.90791933424090698, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale1_score'], 0.89395660507453423, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale2_score'], 0.93010045185439161, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale3_score'], 0.96503534602850938, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif2_score'], 0.72722361912801026, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm3_score'], 0.92425293429958566, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_motion_score'], 4.04982535417, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_motion2_score'], 3.8953518541666665, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_ansnr_score'], 31.2714392708, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_num_score'], 1597314.86733, places=0)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_den_score'], 1597314.95249, places=0)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_num_score'], 397.83378972916671, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_den_score'], 397.83378972916671, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_anpsnr_score'], 41.9266444375, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_scale0_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_scale1_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_scale2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_scale3_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale0_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale1_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale3_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm3_score'], 1.0, places=4)
def test_run_vmaf_fextractor_with_result_store(self):
print 'test on running VMAF feature extractor with result store...'
ref_path = VmafConfig.test_resource_path("yuv", "src01_hrc00_576x324.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "src01_hrc01_576x324.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=0,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':576, 'height':324})
asset_original = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':576, 'height':324})
result_store = FileSystemResultStore(logger=None)
self.fextractor = VmafFeatureExtractor(
[asset, asset_original],
None, fifo_mode=True,
result_store=result_store
)
print ' running for the first time with fresh calculation...'
self.fextractor.run()
result0, result1 = self.fextractor.results
self.assertTrue(os.path.exists(result_store._get_result_file_path(result0)))
self.assertTrue(os.path.exists(result_store._get_result_file_path(result1)))
print ' running for the second time with stored results...'
self.fextractor.run()
results = self.fextractor.results
self.assertAlmostEqual(results[0]['VMAF_feature_vif_score'], 0.4460930625, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_motion_score'], 4.04982535417, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_motion2_score'], 3.8953518541666665, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm2_score'],0.93458780728708746, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_ansnr_score'], 23.5095715208, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_num_score'], 712650.023478, places=0)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_den_score'], 1597314.95249, places=0)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_num_score'], 371.83541406249998, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_den_score'], 397.83378972916671, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_anpsnr_score'], 34.164776875, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_scale0_score'], 0.363420489439, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_scale1_score'], 0.766647542135, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_scale2_score'], 0.862854666902, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif_scale3_score'], 0.915971778036, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale0_score'], 0.90791933424090698, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale1_score'], 0.89395660507453423, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale2_score'], 0.93010045185439161, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale3_score'], 0.96503534602850938, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_vif2_score'], 0.72722361912801026, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm3_score'], 0.92425293429958566, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_motion_score'], 4.04982535417, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_motion2_score'], 3.8953518541666665, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_ansnr_score'], 31.2714392708, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_num_score'], 1597314.86733, places=0)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_den_score'], 1597314.95249, places=0)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_num_score'], 397.83378972916671, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_den_score'], 397.83378972916671, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_anpsnr_score'], 41.9266444375, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_scale0_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_scale1_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_scale2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif_scale3_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale0_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale1_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale3_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_vif2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm3_score'], 1.0, places=4)
def test_run_moment_fextractor(self):
print 'test on running Moment feature extractor...'
ref_path = VmafConfig.test_resource_path("yuv", "src01_hrc00_576x324.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "src01_hrc01_576x324.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=0,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':576, 'height':324})
asset_original = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':576, 'height':324})
self.fextractor = MomentFeatureExtractor(
[asset, asset_original],
None, fifo_mode=True,
result_store=None
)
self.fextractor.run()
results = self.fextractor.results
self.assertAlmostEqual(results[0]['Moment_feature_ref1st_score'], 59.788567297525134, places=4)
self.assertAlmostEqual(results[0]['Moment_feature_ref2nd_score'], 4696.668388042269, places=4)
self.assertAlmostEqual(results[0]['Moment_feature_refvar_score'], 1121.519917231203, places=4)
self.assertAlmostEqual(results[0]['Moment_feature_dis1st_score'], 61.332006624999984, places=4)
self.assertAlmostEqual(results[0]['Moment_feature_dis2nd_score'], 4798.659574041666, places=4)
self.assertAlmostEqual(results[0]['Moment_feature_disvar_score'], 1036.837184348847, places=4)
self.assertAlmostEqual(results[1]['Moment_feature_ref1st_score'], 59.788567297525134, places=4)
self.assertAlmostEqual(results[1]['Moment_feature_ref2nd_score'], 4696.668388042269, places=4)
self.assertAlmostEqual(results[1]['Moment_feature_refvar_score'], 1121.519917231203, places=4)
self.assertAlmostEqual(results[1]['Moment_feature_dis1st_score'], 59.788567297525134, places=4)
self.assertAlmostEqual(results[1]['Moment_feature_dis2nd_score'], 4696.668388042269, places=4)
self.assertAlmostEqual(results[1]['Moment_feature_disvar_score'], 1121.519917231203, places=4)
def test_run_psnr_fextractor(self):
print 'test on running PSNR feature extractor...'
ref_path = VmafConfig.test_resource_path("yuv", "src01_hrc00_576x324.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "src01_hrc01_576x324.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=0,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':576, 'height':324})
asset_original = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':576, 'height':324})
self.fextractor = PsnrFeatureExtractor(
[asset, asset_original],
None, fifo_mode=True,
result_store=None
)
self.fextractor.run()
results = self.fextractor.results
self.assertAlmostEqual(results[0]['PSNR_feature_psnr_score'], 30.755063979166664, places=4)
self.assertAlmostEqual(results[1]['PSNR_feature_psnr_score'], 60.0, places=4)
def test_run_ssim_fextractor(self):
print 'test on running SSIM feature extractor...'
ref_path = VmafConfig.test_resource_path("yuv", "src01_hrc00_576x324.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "src01_hrc01_576x324.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=0,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':576, 'height':324})
asset_original = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':576, 'height':324})
self.fextractor = SsimFeatureExtractor(
[asset, asset_original],
None, fifo_mode=True,
result_store=None
)
self.fextractor.run()
results = self.fextractor.results
self.assertAlmostEqual(results[0]['SSIM_feature_ssim_score'], 0.86322654166666657, places=4)
self.assertAlmostEqual(results[0]['SSIM_feature_ssim_l_score'], 0.9981474583333334, places=4)
self.assertAlmostEqual(results[0]['SSIM_feature_ssim_c_score'], 0.96126793750000006, places=4)
self.assertAlmostEqual(results[0]['SSIM_feature_ssim_s_score'], 0.89773633333333336, places=4)
self.assertAlmostEqual(results[1]['SSIM_feature_ssim_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['SSIM_feature_ssim_l_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['SSIM_feature_ssim_c_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['SSIM_feature_ssim_s_score'], 1.0, places=4)
def test_run_ms_ssim_fextractor(self):
print 'test on running MS-SSIM feature extractor...'
ref_path = VmafConfig.test_resource_path("yuv", "src01_hrc00_576x324.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "src01_hrc01_576x324.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=0,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':576, 'height':324})
asset_original = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':576, 'height':324})
self.fextractor = MsSsimFeatureExtractor(
[asset, asset_original],
None, fifo_mode=True,
result_store=None
)
self.fextractor.run()
results = self.fextractor.results
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_score'], 0.9632498125, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_l_scale0_score'], 0.9981474583333334, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_c_scale0_score'], 0.96126793750000006, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_s_scale0_score'], 0.89773633333333336, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_l_scale1_score'], 0.99899612500000001, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_c_scale1_score'], 0.9857694375, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_s_scale1_score'], 0.941185875, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_l_scale2_score'], 0.99923564583333324, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_c_scale2_score'], 0.997034020833, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_s_scale2_score'], 0.977992145833, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_l_scale3_score'], 0.99929210416666658, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_c_scale3_score'], 0.999588104167, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_s_scale3_score'], 0.99387125, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_l_scale4_score'], 0.99940356249999995, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_c_scale4_score'], 0.999907625, places=4)
self.assertAlmostEqual(results[0]['MS_SSIM_feature_ms_ssim_s_scale4_score'], 0.998222583333, places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_l_scale0_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_c_scale0_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_s_scale0_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_l_scale1_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_c_scale1_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_s_scale1_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_l_scale2_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_c_scale2_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_s_scale2_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_l_scale3_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_c_scale3_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_s_scale3_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_l_scale4_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_c_scale4_score'], 1., places=4)
self.assertAlmostEqual(results[1]['MS_SSIM_feature_ms_ssim_s_scale4_score'], 1., places=4)
def test_run_vmaf_fextractor_checkerboard(self):
print 'test on running VMAF feature extractor on checkerboard...'
ref_path = VmafConfig.test_resource_path("yuv", "checkerboard_1920_1080_10_3_0_0.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "checkerboard_1920_1080_10_3_10_0.yuv")
dis_path2 = VmafConfig.test_resource_path("yuv", "checkerboard_1920_1080_10_3_1_0.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=0,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':1920, 'height':1080})
asset_original = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':1920, 'height':1080})
asset2 = Asset(dataset="test", content_id=0, asset_id=2,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path2,
asset_dict={'width':1920, 'height':1080})
self.fextractor = VmafFeatureExtractor(
[asset, asset_original, asset2],
None, fifo_mode=True,
result_store=None
)
self.fextractor.run()
results = self.fextractor.results
self.assertAlmostEqual(results[0]['VMAF_feature_adm_score'], 0.053996333333333334, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm2_score'], 0.053996333333333334, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale0_score'], 0.23738393128710478, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale1_score'], 0.08524788663335138, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale2_score'], 0.024058909404945077, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale3_score'], 0.018034879735107798, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_motion_score'], 12.554836666666667, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_motion2_score'], 12.554836666666667, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale0_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale1_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale3_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_motion_score'], 12.554836666666667, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_motion2_score'], 12.554836666666667, places=4)
self.assertAlmostEqual(results[2]['VMAF_feature_adm_score'], 0.78533833333333336, places=4)
self.assertAlmostEqual(results[2]['VMAF_feature_adm2_score'], 0.7853384465157921, places=4)
self.assertAlmostEqual(results[2]['VMAF_feature_adm_scale0_score'], 0.72132189911792899, places=4)
self.assertAlmostEqual(results[2]['VMAF_feature_adm_scale1_score'], 0.69259738857522501, places=4)
self.assertAlmostEqual(results[2]['VMAF_feature_adm_scale2_score'], 0.80415911639244586, places=4)
self.assertAlmostEqual(results[2]['VMAF_feature_adm_scale3_score'], 0.82791889676239039, places=4)
self.assertAlmostEqual(results[2]['VMAF_feature_motion_score'], 12.554836666666667, places=4)
self.assertAlmostEqual(results[2]['VMAF_feature_motion2_score'], 12.554836666666667, places=4)
def test_run_vmaf_fextractor_flat(self):
print 'test on running VMAF feature extractor on flat pattern...'
ref_path = VmafConfig.test_resource_path("yuv", "flat_1920_1080_0.yuv")
dis_path = VmafConfig.test_resource_path("yuv", "flat_1920_1080_10.yuv")
asset = Asset(dataset="test", content_id=0, asset_id=0,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=dis_path,
asset_dict={'width':1920, 'height':1080})
asset_original = Asset(dataset="test", content_id=0, asset_id=1,
workdir_root=VmafConfig.workdir_path(),
ref_path=ref_path,
dis_path=ref_path,
asset_dict={'width':1920, 'height':1080})
self.fextractor = VmafFeatureExtractor(
[asset, asset_original],
None, fifo_mode=True,
result_store=None
)
self.fextractor.run()
results = self.fextractor.results
self.assertAlmostEqual(results[0]['VMAF_feature_adm_score'], 1.0, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm2_score'], 1.0, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale0_score'], 1.0, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale1_score'], 1.0, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale2_score'], 1.0, places=4)
self.assertAlmostEqual(results[0]['VMAF_feature_adm_scale3_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale0_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale1_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale2_score'], 1.0, places=4)
self.assertAlmostEqual(results[1]['VMAF_feature_adm_scale3_score'], 1.0, places=4)
