File size: 13,333 Bytes
a103028 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 |
#!/usr/local/bin/python
import argparse
from common import *
class Depth:
def __init__(self, preset = None, architecture = None):
self.preset = preset
self.architecture = architecture
def command(self, background_image, depth_image):
AssertFileExists(background_image)
command = "depth-estimation " + " --input=" + background_image + " --output=" + depth_image
if self.architecture is not None:
command = command + " --architecture=" + self.architecture
if self.preset is not None:
command = command + " --preset=" + self.preset
return command
class Slate:
def __init__(self, architecture = None, slate_begin = None, slate_end = None):
self.architecture = architecture
self.slate_begin = slate_begin
self.slate_end = slate_end
def command(self, heartbeats, wide_video, wide_slate, tele_video = None, tele_slate = None):
AssertFileExists(wide_video)
command = "background-extraction " + " --input-video=" + wide_video + " --output-slate=" + wide_slate
if self.architecture == "heartbeats" and FileExists(heartbeats):
command = command + " --input-heartbeats=" + heartbeats
if FileExists(tele_video) and (tele_slate is not None):
command = command + " --input-video-multicam-tele=" + tele_video
command = command + " --output-slate-multicam-tele=" + tele_slate
return command
class Matte:
def __init__(self, architecture, model, num_frames = 0):
self.architecture = architecture
self.model = model
self.num_frames = num_frames
def command(self, input_video, output_video, background_image = None):
command = "human-segmentation --input=" + input_video + " --output=" + output_video
if FileExists(background_image):
command = command + " --slate=" + background_image
if self.architecture != None:
command = command + " --architecture=" + self.architecture
if self.model != None:
command = command + " --model=" + self.model
if self.num_frames is not None and self.num_frames > 0:
command = command + " --num-frames=" + str(self.num_frames)
return command
class MatteAnalysis:
def __init__(self):
pass
def command(self, input_video, input_matte, input_heartbeats, transparent_video=None):
command = "python matte-analysis.py -iv {:s} -i {:s} -hb {:s}".format(input_video, input_matte, input_heartbeats)
if transparent_video is not None:
command += " -tv {:s}".format(transparent_video)
return command
## ----
def GetCameraIdInDir(sequence_dir, content_name):
wide_file_path = VideoFilePath(sequence_dir, content_name, "WA")
tele_file_path = VideoFilePath(sequence_dir, content_name, "TP")
if FileExists(wide_file_path) and FileExists(tele_file_path):
return ["WA", "TP"]
wide_file_path = VideoFilePath(sequence_dir, content_name, "1X")
tele_file_path = VideoFilePath(sequence_dir, content_name, "2X")
if FileExists(wide_file_path) and FileExists(tele_file_path):
return ["1X", "2X"]
for camera_id in ["WA", "1X", "TP", "2X", "TD", "LC", "FF", "UW"]:
video_file_path = VideoFilePath(sequence_dir, content_name, camera_id)
if FileExists( video_file_path ):
return [camera_id]
return []
def MatteFilePath(sequence_dir, sequence_name, camera_id, file_must_exist = True):
segmentation_video = sequence_dir + sequence_name + "-" + camera_id + "-SM.MP4"
if file_must_exist == False:
return segmentation_video
if FileExists(segmentation_video) and file_must_exist == True:
return segmentation_video
return ""
def DepthFilePath(sequence_dir, sequence_name, camera_id, file_must_exist = True):
background_image_depth = sequence_dir + sequence_name + "-" + camera_id + "-DM.PNG"
if file_must_exist == False:
return background_image_depth
if FileExists(background_image_depth) and file_must_exist == True:
return background_image_depth
return ""
def SlateFilePath(sequence_dir, sequence_name, camera_id, file_must_exist = True):
background_image = sequence_dir + sequence_name + "-" + camera_id + "-BG.PNG"
if file_must_exist == False:
return background_image
if FileExists(background_image) and file_must_exist == True:
return background_image
background_image = sequence_dir + sequence_name + "-" + camera_id + "-BG.JPG"
if FileExists(background_image) and file_must_exist == True:
return background_image
return ""
def VideoFilePath(sequence_dir, sequence_name, camera_id):
video_file_path = sequence_dir + sequence_name + "-" + camera_id + ".MP4"
if FileExists(video_file_path):
return video_file_path
video_file_path = sequence_dir + sequence_name + "-" + camera_id + ".MOV"
if FileExists(video_file_path):
return video_file_path
return ""
def HbeatFilePath(sequence_dir, sequence_name):
input_heartbeats = sequence_dir + sequence_name + "-HB.ZIP"
if FileExists(input_heartbeats):
return input_heartbeats
input_heartbeats = sequence_dir + sequence_name + "-HB.TXT"
if FileExists(input_heartbeats):
return input_heartbeats
return ""
## ----
def sanitize_args(args):
AssertDirExists(args.content)
args.content = GetFullPath(args.content)
if args.output_dir is not None:
args.output_dir = GetFullPath(args.output_dir)
return args
def GetTestParams(content_dir, experiment_name):
content_name = GetDirBaseName(content_dir)
output_file_name = content_name
if experiment_name != None:
output_file_name = output_file_name + "-" + experiment_name
if args.output_dir == "content":
output_dir = content_dir
elif args.output_dir is not None:
output_dir = args.output_dir
MakeDirRecursive(output_dir)
return output_file_name, content_name, output_dir
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument('-c', '--content', type=str, default='', required=True, help='Content folder')
parser.add_argument('-o', '--output-dir', type=str, default="./", required=False, help='Output (results) folder')
parser.add_argument('-n', '--name', type=str, default=None, required=False, help='Name of experiment')
parser.add_argument('-m', '--modules', type=str, default=None, nargs='+', required=True, choices=["matte", "slate", "depth"], help='Module to test')
# matte params
parser.add_argument('--matte-num-frames', type=int, default=None, help='Number of frames to process for matting')
parser.add_argument('--matte-architecture', type=str, default=None, required=False, choices=["washington", "washington-base", "temporal"], help='Matte architecture to use.')
parser.add_argument('--matte-model', type=str, default=None, required=False, choices=["resnet101", "resnet50", "mobilenetv2", "mobilenetv3"], help='Matte model')
# slate params
parser.add_argument('--slate-architecture', type=str, default="heartbeats", required=False, choices=["dlib", "heartbeats"], help='Slate architecture.')
# depth params
parser.add_argument('--depth-architecture', type=str, default=None, required=False, choices=["midas-v1-only", "multires-midas", "multires-structrl", "multires-leres"], help='Depth estimation architecture.')
parser.add_argument('--depth-preset', type=str, default=None, required=False, choices=["precise", "fast", "veryfast"], help='Depth estimation architecture.')
parser.add_argument('--log-timestamp', type=int, default=DefaultLogTimestamp(), help=LogTimestampHelp())
parser.add_argument('--log-level', type=str, default=DefaultLogLevel(), help=LogLevelHelp())
# matte analysis params
parser.add_argument('--transparent-video', type=bool, default=None, help="Specify if transparent video be written to output")
args = parser.parse_args()
args = sanitize_args(args)
InitLog(args.log_level, args.log_timestamp)
# # printHeader(GetRealPath(__file__))
content_dirs = GetAllSubDirs(args.content)
matte = Matte(args.matte_architecture, args.matte_model, args.matte_num_frames)
slate = Slate(args.slate_architecture)
depth = Depth(args.depth_preset, args.depth_architecture)
# Slate
command_list = []
if "slate" in args.modules:
for content_dir in content_dirs:
output_file_name, content_name, output_dir = GetTestParams(content_dir, args.name)
camera_ids = GetCameraIdInDir(content_dir, content_name)
if len(camera_ids) >= 1:
hbeat_path = HbeatFilePath(content_dir, content_name)
video_path = VideoFilePath(content_dir, content_name, camera_ids[0])
if FileExists(video_path) and FileExists(hbeat_path):
command = ""
if len(camera_ids) == 2:
# Multicam case
video_path_multicam_tele = VideoFilePath(content_dir, content_name, camera_ids[1])
if FileExists(video_path_multicam_tele):
command = slate.command(hbeat_path,
video_path, SlateFilePath(output_dir, output_file_name, camera_ids[0], False),
video_path_multicam_tele, SlateFilePath(output_dir, output_file_name, camera_ids[1], False))
else:
# Singlecam case
command = slate.command(hbeat_path, video_path,
SlateFilePath(output_dir, output_file_name, camera_ids[0], False))
if command != "":
command_list.append({ 'command': command })
else:
log.warning("Could not create command for " + content_dir)
else:
log.warning("Could not find any valid camera id in directory " + content_dir)
ExecuteCommandList(command_list, GetNumberOfVirtualCores())
# Depth
command_list = []
if "depth" in args.modules:
for content_dir in content_dirs:
output_file_name, content_name, output_dir = GetTestParams(content_dir, args.name)
for camera_id in GetCameraIdInDir(content_dir, content_name):
slate_path = SlateFilePath(content_dir, content_name, camera_id, True)
if FileExists(slate_path):
command = depth.command(slate_path, DepthFilePath(output_dir, output_file_name, camera_id, False))
command_list.append({ 'command': command })
ExecuteCommandList(command_list, 1)
# Matte
command_list = []
if "matte" in args.modules:
for content_dir in content_dirs:
output_file_name, content_name, output_dir = GetTestParams(content_dir, args.name)
for camera_id in GetCameraIdInDir(content_dir, content_name):
video_path = VideoFilePath(content_dir, content_name, camera_id)
if FileExists(video_path):
command = matte.command(video_path, MatteFilePath(output_dir, output_file_name, camera_id, False),
SlateFilePath(content_dir, content_name, camera_id, True))
command_list.append({ 'command': command })
ExecuteCommandList(command_list, 1)
# printFooter(GetRealPath(__file__), 0)
# def TransparentVideoPath(sequence_dir, sequence_name, camera_id, file_must_exist = True):
# transparent_video = sequence_dir + sequence_name + "-" + camera_id + "-MA.MP4"
# if file_must_exist == False:
# return transparent_video
# if FileExists(transparent_video) and file_must_exist == True:
# return transparent_video
# return ""
# TODO(Isam) - Why do you need matte analyzer in regression test? - Plz discuss with me
# # Matte Analysis
# command_list = []
# valid_camera_ids = ['WA', '1X']
# if "matte-analysis" in args.tests:
# for content_dir in content_dirs:
# output_file_name, content_name, output_dir = GetTestParams(content_dir, args.experiment)
# for camera_id in GetAllCameraIds():
# if camera_id not in valid_camera_ids:
# continue
# video_file_path = VideoFilePath(content_dir, content_name, camera_id)
# if args.transparent_video:
# transparent_video_path = TransparentVideoPath(output_dir, output_file_name, camera_id, False)
# if FileExists(video_file_path):
# command = matte_analysis.command(video_file_path, MatteFilePath(content_dir, content_name, camera_id, True), \
# HbeatFilePath(content_dir, content_name), transparent_video=transparent_video_path)
# command_list.append({'command': command,
# 'stdout': LogFilePath(args.log, output_dir, output_file_name, camera_id, False)}) |