Skip to content
Snippets Groups Projects
Commit b68afbd1 authored by Amir Mohammadi's avatar Amir Mohammadi
Browse files

refactor fusion script, fix bugs

parent f643a2c2
Branches
Tags
1 merge request!2First Alpha release
Hide whitespace changes
Inline Side-by-side
bob/fusion/base/script/bob_fuse.py
+ 152
− 121
View file @ b68afbd1
...@@ -12,138 +12,169 @@ from bob.measure.load import load_score, dump_score ...@@ -12,138 +12,169 @@ from bob.measure.load import load_score, dump_score
from bob.bio.base import utils from bob.bio.base import utils
from ..tools import parse_arguments, write_info, get_gza_from_lines_list, \ from ..tools import parse_arguments, write_info, get_gza_from_lines_list, \
check_consistency, get_scores, remove_nan, get_score_lines, get_2negatives_1positive check_consistency, get_scores, remove_nan, get_score_lines, get_2negatives_1positive
import bob.core import bob.core
logger = bob.core.log.setup("bob.fusion.base") logger = bob.core.log.setup("bob.fusion.base")
def save_fused_scores(save_path, fused_scores, score_lines):
score_lines['score'] = fused_scores
gen, zei, atk, _, _, _ = get_2negatives_1positive(score_lines)
create_directories_safe(os.path.dirname(save_path))
dump_score(save_path, score_lines)
dump_score(save_path + '-licit', np.append(gen, zei))
dump_score(save_path + '-spoof', np.append(gen, atk))
def routine_fusion(algorithm, model_file,
scores_train_lines, scores_train, train_neg, train_pos, fused_train_file,
scores_dev_lines=None, scores_dev=None, dev_neg=None, dev_pos=None, fused_dev_file=None,
scores_eval_lines=None, scores_eval=None, fused_eval_file=None,
force=False, min_file_size=1000,
):
# train the preprocessors
train_scores = np.vstack((train_neg, train_pos))
neg_len = train_neg.shape[0]
y = np.zeros((train_scores.shape[0],), dtype='bool')
y[neg_len:] = True
algorithm.train_preprocessors(train_scores, y)
# preprocess data
scores_train = algorithm.preprocess(scores_train)
train_neg, train_pos = algorithm.preprocess(train_neg), algorithm.preprocess(train_pos)
if scores_dev is not None:
scores_dev = algorithm.preprocess(scores_dev)
dev_neg, dev_pos = algorithm.preprocess(dev_neg), algorithm.preprocess(dev_pos)
if scores_eval is not None:
scores_eval = algorithm.preprocess(scores_eval)
# train the model
if utils.check_file(model_file, force, min_file_size):
logger.info(
"model '%s' already exists.", model_file)
algorithm = algorithm.load(model_file)
else:
algorithm.train(train_neg, train_pos, dev_neg, dev_pos)
algorithm.save(model_file)
# fuse the scores (train)
if utils.check_file(fused_train_file, force, min_file_size):
logger.info(
"score file '%s' already exists.", fused_train_file)
else:
fused_scores_train = algorithm.fuse(scores_train)
save_fused_scores(fused_train_file, fused_scores_train, scores_train_lines)
if scores_dev is not None:
# fuse the scores (dev)
if utils.check_file(fused_dev_file, force, min_file_size):
logger.info(
"score file '%s' already exists.", fused_dev_file)
else:
fused_scores_dev = algorithm.fuse(scores_dev)
save_fused_scores(fused_dev_file, fused_scores_dev, scores_dev_lines)
if scores_eval is not None:
# fuse the scores (eval)
if utils.check_file(fused_eval_file, force, min_file_size):
logger.info(
"score file '%s' already exists.", fused_eval_file)
else:
fused_scores_eval = algorithm.fuse(scores_eval)
save_fused_scores(fused_eval_file, fused_scores_eval, scores_eval_lines)
def fuse(args, command_line_parameters): def fuse(args, command_line_parameters):
"""Do the actual fusion.""" """Do the actual fusion."""
algorithm = args.algorithm algorithm = args.algorithm
write_info(args, command_line_parameters) write_info(args, command_line_parameters)
# load the scores # load the scores
score_lines_list_train = [load_score(path, ncolumns=args.score_type) score_lines_list_train = [load_score(path, ncolumns=args.score_type)
for path in args.train_files] for path in args.train_files]
if args.dev_files:
score_lines_list_dev = [load_score(path, ncolumns=args.score_type)
for path in args.dev_files]
if args.eval_files:
score_lines_list_eval = [load_score(path, ncolumns=args.score_type)
for path in args.eval_files]
# genuine, zero effort impostor, and attack list of
# train, development and evaluation data.
idx1, gen_lt, zei_lt, atk_lt = get_gza_from_lines_list(score_lines_list_train)
if args.dev_files:
_, gen_ld, zei_ld, atk_ld = get_gza_from_lines_list(score_lines_list_dev)
if args.eval_files:
_, gen_le, zei_le, atk_le = get_gza_from_lines_list(score_lines_list_eval)
# check if score lines are consistent
if not args.skip_check:
logger.info('Checking the training files for consistency ...')
check_consistency(gen_lt, zei_lt, atk_lt)
if args.dev_files: if args.dev_files:
logger.info('Checking the development files for consistency ...') score_lines_list_dev = [load_score(path, ncolumns=args.score_type)
check_consistency(gen_ld, zei_ld, atk_ld) for path in args.dev_files]
if args.eval_files: if args.eval_files:
logger.info('Checking the evaluation files for consistency ...') score_lines_list_eval = [load_score(path, ncolumns=args.score_type)
check_consistency(gen_le, zei_le, atk_le) for path in args.eval_files]
scores_train = get_scores(gen_lt, zei_lt, atk_lt) # genuine, zero effort impostor, and attack list of
train_neg = get_scores(zei_lt, atk_lt) # train, development and evaluation data.
train_pos = get_scores(gen_lt) idx1, gen_lt, zei_lt, atk_lt = get_gza_from_lines_list(score_lines_list_train)
if args.dev_files: if args.dev_files:
scores_dev = get_scores(gen_ld, zei_ld, atk_ld) _, gen_ld, zei_ld, atk_ld = get_gza_from_lines_list(score_lines_list_dev)
scores_dev_lines = get_score_lines(gen_ld[0:1], zei_ld[0:1], atk_ld[0:1]) if args.eval_files:
dev_neg = get_scores(zei_ld, atk_ld) _, gen_le, zei_le, atk_le = get_gza_from_lines_list(score_lines_list_eval)
dev_pos = get_scores(gen_ld)
else: # check if score lines are consistent
dev_neg, dev_pos = None, None if not args.skip_check:
if args.eval_files: logger.info('Checking the training files for consistency ...')
scores_eval = get_scores(gen_le, zei_le, atk_le) check_consistency(gen_lt, zei_lt, atk_lt)
scores_eval_lines = get_score_lines(gen_le[0:1], zei_le[0:1], atk_le[0:1]) if args.dev_files:
logger.info('Checking the development files for consistency ...')
# check for nan values check_consistency(gen_ld, zei_ld, atk_ld)
found_nan = False if args.eval_files:
found_nan, _, scores_train = remove_nan(scores_train, found_nan) logger.info('Checking the evaluation files for consistency ...')
found_nan, _, train_neg = remove_nan(train_neg, found_nan) check_consistency(gen_le, zei_le, atk_le)
found_nan, _, train_pos = remove_nan(train_pos, found_nan)
if args.dev_files: scores_train = get_scores(gen_lt, zei_lt, atk_lt)
found_nan, nan_dev, scores_dev = remove_nan(scores_dev, found_nan) scores_train_lines = get_score_lines(gen_lt[0:1], zei_lt[0:1], atk_lt[0:1])
found_nan, _, dev_neg = remove_nan(dev_neg, found_nan) train_neg = get_scores(zei_lt, atk_lt)
found_nan, _, dev_pos = remove_nan(dev_pos, found_nan) train_pos = get_scores(gen_lt)
if args.eval_files: if args.dev_files:
found_nan, nan_eval, scores_eval = remove_nan(scores_eval, found_nan) scores_dev = get_scores(gen_ld, zei_ld, atk_ld)
scores_dev_lines = get_score_lines(gen_ld[0:1], zei_ld[0:1], atk_ld[0:1])
if found_nan: dev_neg = get_scores(zei_ld, atk_ld)
logger.warn('Some nan values were removed.') dev_pos = get_scores(gen_ld)
# train the preprocessors
algorithm.train_preprocessors(scores_train)
# preprocess data
train_neg, train_pos = algorithm.preprocess(train_neg), algorithm.preprocess(train_pos)
if args.dev_files:
scores_dev = algorithm.preprocess(scores_dev)
dev_neg, dev_pos = algorithm.preprocess(dev_neg), algorithm.preprocess(dev_pos)
if args.eval_files:
scores_eval = algorithm.preprocess(scores_eval)
# train the model
if utils.check_file(args.model_file, args.force, 1000):
logger.info(
"- Fusion: model '%s' already exists.", args.model_file)
algorithm = algorithm.load(args.model_file)
else:
algorithm.train(train_neg, train_pos, dev_neg, dev_pos)
algorithm.save(args.model_file)
# fuse the scores (dev)
if utils.check_file(args.fused_dev_file, args.force, 1000):
logger.info(
"- Fusion: scores '%s' already exists.", args.fused_dev_file)
elif args.dev_files:
fused_scores_dev = algorithm.fuse(scores_dev)
score_lines = scores_dev_lines[~nan_dev]
score_lines['score'] = fused_scores_dev
gen, zei, atk, _, _, _ = get_2negatives_1positive(score_lines)
create_directories_safe(os.path.dirname(args.fused_dev_file))
dump_score(args.fused_dev_file, score_lines)
dump_score(args.fused_dev_file + '-licit', np.append(gen, zei))
dump_score(args.fused_dev_file + '-spoof', np.append(gen, atk))
# fuse the scores (eval)
if args.eval_files:
if utils.check_file(args.fused_eval_file, args.force, 1000):
logger.info(
"- Fusion: scores '%s' already exists.", args.fused_eval_file)
else: else:
fused_scores_eval = algorithm.fuse(scores_eval) scores_dev, scores_dev_lines, dev_neg, dev_pos = None, None, None, None
score_lines = scores_eval_lines[~nan_eval] if args.eval_files:
score_lines['score'] = fused_scores_eval scores_eval = get_scores(gen_le, zei_le, atk_le)
gen, zei, atk, _, _, _ = get_2negatives_1positive(score_lines) scores_eval_lines = get_score_lines(gen_le[0:1], zei_le[0:1], atk_le[0:1])
create_directories_safe(os.path.dirname(args.fused_eval_file)) else:
dump_score(args.fused_eval_file, score_lines) scores_eval, scores_eval_lines = None, None
dump_score(args.fused_eval_file + '-licit', np.append(gen, zei))
dump_score(args.fused_eval_file + '-spoof', np.append(gen, atk)) # check for nan values
found_nan = False
found_nan, nan_train, scores_train = remove_nan(scores_train, found_nan)
scores_train_lines = scores_train_lines[~nan_train]
found_nan, _, train_neg = remove_nan(train_neg, found_nan)
found_nan, _, train_pos = remove_nan(train_pos, found_nan)
if args.dev_files:
found_nan, nan_dev, scores_dev = remove_nan(scores_dev, found_nan)
scores_dev_lines = scores_dev_lines[~nan_dev]
found_nan, _, dev_neg = remove_nan(dev_neg, found_nan)
found_nan, _, dev_pos = remove_nan(dev_pos, found_nan)
if args.eval_files:
found_nan, nan_eval, scores_eval = remove_nan(scores_eval, found_nan)
scores_eval_lines = scores_eval_lines[~nan_eval]
if found_nan:
logger.warn('Some nan values were removed.')
routine_fusion(algorithm, args.model_file,
scores_train_lines, scores_train, train_neg, train_pos, args.fused_train_file,
scores_dev_lines, scores_dev, dev_neg, dev_pos, args.fused_dev_file,
scores_eval_lines, scores_eval, args.fused_eval_file,
args.force)
def main(command_line_parameters=None): def main(command_line_parameters=None):
"""Executes the main function""" """Executes the main function"""
try: try:
# do the command line parsing # do the command line parsing
args = parse_arguments(command_line_parameters) args = parse_arguments(command_line_parameters)
# perform face verification test # perform face verification test
fuse(args, command_line_parameters) fuse(args, command_line_parameters)
except Exception as e: except Exception as e:
# track any exceptions as error logs (i.e., to get a time stamp) # track any exceptions as error logs (i.e., to get a time stamp)
logger.error("During the execution, an exception was raised: %s" % e) logger.error("During the execution, an exception was raised: %s" % e)
raise raise
if __name__ == "__main__": if __name__ == "__main__":
main() main()
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment