def test_big_pipeline():
pytest.importorskip('pygraphviz')
filename = join(get_test_data_path(), 'video', 'obama_speech.mp4')
video = VideoStim(filename)
visual_nodes = [(FrameSamplingFilter(every=15), [
(TesseractConverter(), [LengthExtractor()]),
VibranceExtractor(),
'BrightnessExtractor',
])]
audio_nodes = [(VideoToAudioConverter(),
[WitTranscriptionConverter(),
'LengthExtractor'], 'video_to_audio')]
graph = Graph()
graph.add_nodes(visual_nodes)
graph.add_nodes(audio_nodes)
results = graph.run(video, merge=False)
result = merge_results(results, format='wide', extractor_names='multi')
# Test that pygraphviz outputs a file
drawfile = next(tempfile._get_candidate_names())
graph.draw(drawfile)
assert exists(drawfile)
os.remove(drawfile)
assert ('LengthExtractor', 'text_length') in result.columns
assert ('VibranceExtractor', 'vibrance') in result.columns
# assert not result[('onset', '')].isnull().any()
assert 'text[negotiations]' in result['stim_name'].values
assert 'frame[90]' in result['stim_name'].values
def test_vibrance_extractor():
stim = ImageStim(join(IMAGE_DIR, 'apple.jpg'), onset=4.2, duration=1)
result = VibranceExtractor().transform(stim).to_df()
color = result['vibrance'][0]
assert np.isclose(color, 1370.65482988, 1e-5)
assert result['onset'][0] == 4.2
assert result['duration'][0] == 1
def test_vibrance_extractor():
image_dir = join(get_test_data_path(), 'image')
stim = ImageStim(join(image_dir, 'apple.jpg'), onset=4.2, duration=1)
result = VibranceExtractor().transform(stim).to_df()
color = result['vibrance'][0]
assert np.isclose(color, 1370.65482988, 1e-5)
assert result['onset'][0] == 4.2
assert result['duration'][0] == 1
def test_small_pipeline2():
filename = join(get_test_data_path(), 'image', 'button.jpg')
nodes = [BrightnessExtractor(), VibranceExtractor()]
graph = Graph(nodes)
result = list(graph.run([filename], merge=False))
history = result[0].history.to_df()
assert history.shape == (1, 8)
result = merge_results(result, format='wide', extractor_names='multi')
assert ('BrightnessExtractor', 'brightness') in result.columns
brightness = result[('BrightnessExtractor', 'brightness')].values[0]
vibrance = result[('VibranceExtractor', 'vibrance')].values[0]
assert_almost_equal(brightness, 0.746965, 5)
assert ('VibranceExtractor', 'vibrance') in result.columns
assert_almost_equal(vibrance, 841.577274, 5)
def test_merge_extractor_results_by_features():
np.random.seed(100)
image_dir = join(get_test_data_path(), 'image')
stim = ImageStim(join(image_dir, 'apple.jpg'))
# Merge results for static Stims (no onsets)
extractors = [BrightnessExtractor(), VibranceExtractor()]
results = [e.transform(stim) for e in extractors]
df = ExtractorResult.merge_features(results)
de = DummyExtractor()
de_names = ['Extractor1', 'Extractor2', 'Extractor3']
results = [de.transform(stim, name) for name in de_names]
df = ExtractorResult.merge_features(results)
assert df.shape == (177, 14)
assert df.columns.levels[1].unique().tolist() == ['duration', 0, 1, 2, '']
cols = cols = ['onset', 'class', 'filename', 'history', 'stim']
assert df.columns.levels[0].unique().tolist() == de_names + cols
def test_big_pipeline():
filename = join(get_test_data_path(), 'video', 'obama_speech.mp4')
video = VideoStim(filename)
visual_nodes = [(FrameSamplingConverter(every=15), [
(TesseractConverter(), [LengthExtractor()]),
VibranceExtractor(), 'BrightnessExtractor',
])]
audio_nodes = [(VideoToAudioConverter(), [
WitTranscriptionConverter(), 'LengthExtractor'],
'video_to_audio')]
graph = Graph()
graph.add_nodes(visual_nodes)
graph.add_nodes(audio_nodes)
result = graph.run(video)
assert ('LengthExtractor', 'text_length') in result.columns
assert ('VibranceExtractor', 'vibrance') in result.columns
# assert not result[('onset', '')].isnull().any()
assert 'text[negotiations]' in result['stim'].values
assert 'frame[90]' in result['stim'].values
def extract_visual_features(video_file):
"""
This function extracts luminance, vibrance, saliency, and sharpness from the frames of a video
using the pliers library. If you use this function, please cite the pliers library directly:
https://github.com/PsychoinformaticsLab/pliers#how-to-cite
Parameters
----------
video_file: str
Path to video file to analyze.
Returns
-------
low_level_video_df: DataFrame
Pandas dataframe with a column per low-level feature.py (index is time).
"""
# extract video luminance
print('Extracting brightness...')
from pliers.extractors import BrightnessExtractor
brightext = BrightnessExtractor()
brightres = brightext.transform(video_file)
brightres_df = pd.DataFrame(columns=brightres[0].to_df().columns)
for a, ob in enumerate(brightres):
t = ob.to_df()
t['order'] = a
brightres_df = brightres_df.append(t, ignore_index=True)
# extract saliency
print('Extracting saliency...')
from pliers.extractors import SaliencyExtractor
salext = SaliencyExtractor()
salres = salext.transform(video_file)
salres_df = pd.DataFrame(columns=salres[0].to_df().columns)
for a, ob in enumerate(salres):
t = ob.to_df()
t['order'] = a
salres_df = salres_df.append(t, ignore_index=True)
# extract sharpness
print('Extracting sharpness...')
from pliers.extractors import SharpnessExtractor
sharpext = SharpnessExtractor()
sharpres = sharpext.transform(video_file)
sharpres_df = pd.DataFrame(columns=sharpres[0].to_df().columns)
for a, ob in enumerate(sharpres):
t = ob.to_df()
t['order'] = a
sharpres_df = sharpres_df.append(t, ignore_index=True)
# extract vibrance
print('Extracting vibrance...')
from pliers.extractors import VibranceExtractor
vibext = VibranceExtractor()
vibres = vibext.transform(video_file)
vibres_df = pd.DataFrame(columns=vibres[0].to_df().columns)
for a, ob in enumerate(vibres):
t = ob.to_df()
t['order'] = a
vibres_df = vibres_df.append(t, ignore_index=True)
# combine into 1 dataframe
print('Combining data...')
low_level_video_df = brightres_df.merge(salres_df[salres_df.columns[4:]],
left_index=True,
right_index=True)
low_level_video_df = low_level_video_df.merge(
sharpres_df[sharpres_df.columns[4:]],
left_index=True,
right_index=True)
low_level_video_df = low_level_video_df.merge(
vibres_df[vibres_df.columns[4:]], left_index=True, right_index=True)
low_level_video_df['onset_ms'] = low_level_video_df['onset'] * 1000
low_level_video_df.index = pd.to_datetime(low_level_video_df['onset_ms'],
unit='ms')
low_level_video_df = low_level_video_df.drop(
['max_saliency', 'max_y', 'max_x', 'onset', 'object_id', 'order'],
axis=1)
low_level_video_df.index.name = None
print('Visual feature extraction complete.')
return low_level_video_df