def main():
"""
Entry point of the script for using a trained model for predicting videos.
i.e: python predict.py -w data/res/2018-09-26-02-30/lipnet_065_1.96.hdf5 -v data/dataset_eval
"""
print(r'''
__ __ ______ __ __ ______ ______
/\ \ /\ \ /\ == \ /\ "-.\ \ /\ ___\ /\__ _\
\ \ \____ \ \ \ \ \ _-/ \ \ \-. \ \ \ __\ \/_/\ \/
\ \_____\ \ \_\ \ \_\ \ \_\\"\_\ \ \_____\ \ \_\
\/_____/ \/_/ \/_/ \/_/ \/_/ \/_____/ \/_/
implemented by Omar Salinas
''')
ap = argparse.ArgumentParser()
ap.add_argument('-v',
'--video-path',
required=True,
help='Path to video file or batch directory to analize')
ap.add_argument('-w',
'--weights-path',
required=True,
help='Path to .hdf5 trained weights file')
default_predictor = os.path.join(__file__, '..', 'data', 'predictors',
'shape_predictor_68_face_landmarks.dat')
ap.add_argument("-pp",
"--predictor-path",
required=False,
help="(Optional) Path to the predictor .dat file",
default=default_predictor)
args = vars(ap.parse_args())
weights = os.path.realpath(args['weights_path'])
video = os.path.realpath(args['video_path'])
predictor_path = os.path.realpath(args["predictor_path"])
if not is_file(weights) or get_file_extension(weights) != '.hdf5':
print(Fore.RED + '\nERROR: Trained weights path is not a valid file')
return
if not is_file(video) and not is_dir(video):
print(
Fore.RED +
'\nERROR: Path does not point to a video file nor to a directory')
return
if not is_file(
predictor_path) or get_file_extension(predictor_path) != '.dat':
print(Fore.RED + '\nERROR: Predictor path is not a valid file')
return
fcount = video_to_frames(os.path.join(video, 'p.mpg'))
config = PredictConfig(weights, video, predictor_path, fcount)
predict(config)
def get_video_data(path: str, detector, predictor) -> np.ndarray:
if get_file_extension(path) == '.mpg':
data = extract_video_data(path, detector, predictor, False)
return reshape_and_normalize_video_data(
data) if data is not None else None
else:
return get_video_data_from_file(path)
def main():
print(r'''
__ __ ______ __ __ ______ __ __ ______
/\ \ /\ \ /\ == \ /\ "-.\ \ /\ ___\ /\_\_\_\ /\__ _\
\ \ \____ \ \ \ \ \ _-/ \ \ \-. \ \ \ __\ \/_/\_\/_ \/_/\ \/
\ \_____\ \ \_\ \ \_\ \ \_\\"\_\ \ \_____\ /\_\/\_\ \ \_\
\/_____/ \/_/ \/_/ \/_/ \/_/ \/_____/ \/_/\/_/ \/_/
''')
ap = argparse.ArgumentParser()
ap.add_argument("-v", "--videos-path", required=True, help="Path to videos directory")
ap.add_argument("-o", "--output-path", required=True, help="Path for the extracted frames")
default_predictor = os.path.join(__file__, '..', '..', 'data', 'predictors', 'shape_predictor_68_face_landmarks.dat')
ap.add_argument("-pp", "--predictor-path", required=False, help="(Optional) Path to the predictor .dat file", default=default_predictor)
ap.add_argument("-p", "--pattern", required=False, help="(Optional) File name pattern to match", default='*.mpg')
ap.add_argument("-fv", "--first-video", required=False, help="(Optional) First video index extracted in each speaker (inclusive)", type=int, default=0)
ap.add_argument("-lv", "--last-video", required=False, help="(Optional) Last video index extracted in each speaker (exclusive)", type=int, default=1000)
args = vars(ap.parse_args())
videos_path = os.path.realpath(args["videos_path"])
output_path = os.path.realpath(args["output_path"])
pattern = args["pattern"]
first_video = args["first_video"]
last_video = args["last_video"]
predictor_path = os.path.realpath(args["predictor_path"])
if not is_dir(videos_path):
print(Fore.RED + 'ERROR: Invalid path to videos directory')
return
if not isinstance(output_path, str):
print(Fore.RED + 'ERROR: Invalid path to output directory')
return
if not is_file(predictor_path) or get_file_extension(predictor_path) != '.dat':
print(Fore.RED + '\nERROR: Predictor path is not a valid file')
return
if not isinstance(first_video, int) or first_video < 0:
print(Fore.RED + '\nERROR: The first video index must be a valid positive integer')
return
if not isinstance(last_video, int) or last_video < first_video:
print(Fore.RED + '\nERROR: The last video index must be a valid positive integer greater than the first video index')
return
extract_to_npy(videos_path, output_path, predictor_path, pattern, first_video, last_video)
def get_entire_video_data(path: str) -> np.ndarray:
if get_file_extension(path) == '.mpg':
return np.swapaxes(skvideo.io.vread(path), 1, 2)
else:
return get_video_data_from_file(path)