/
ubm_adaptation.py
162 lines (135 loc) · 5.72 KB
/
ubm_adaptation.py
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
import os
import gzip
import cPickle
import argparse
import numpy as np
import sys
import progressbar
#import pyvole
import puhma_common as pc
from encoding import *
import evaluate
def parserArguments(parser):
ubm_group = parser.add_argument_group('adaptation options')
ubm_group.add_argument('--load_ubm',\
help='filepath to pkl.gz file which contains the ubm-gmm')
ubm_group.add_argument('--load_scores',\
help='load score-file (pkl.gz)')
ubm_group.add_argument('--encoding',choices =['fisher',
'supervector',
'vlad'],\
help='different encoding schemes of the gmm')
ubm_group.add_argument('--normalize', nargs='*', default=[],
choices=['l2g','l2c','ssr'],
help='normalization options')
ubm_group.add_argument('-r', '--relevance',type=int, default=16,\
help=('relevance factor for mixing feature vectors'\
' with UBM'))
parser.add_argument('--update', default='wmc',\
help='what to update w. GMM, w:weights, m:means, c:covars')
parser.add_argument('--no_eval', action='store_true',
help='dont evaluate skip evaluation and just write'
' encodings ')
return parser
def loadUBM(ubm_file):
with gzip.open(ubm_file, 'rb') as f:
ubm_gmm = cPickle.load(f)
if hasattr(ubm_gmm, 'weights_') and ubm_gmm.weights_.ndim > 1:
ubm_gmm.weights_ = ubm_gmm.weights_.flatten()
return ubm_gmm
if __name__ == '__main__':
# import stacktracer
# stacktracer.trace_start("trace.html",interval=5,auto=True)
parser = argparse.ArgumentParser(description="UBM Adaption")
parser = pc.commonArguments(parser)
parser = parserArguments(parser)
args = parser.parse_args()
if not args.labelfile or not args.inputfolder or not args.outputfolder:
print('WARNING: no labelfile or no inputfolder'
' or no outputfolder specified')
if args.outputfolder and not os.path.exists(args.outputfolder):
pc.mkdir_p(args.outputfolder)
if not args.load_ubm:
raise argparse.ArgumentTypeError('no gmm to load')
#####
# UBM-creation / loading
print 'load gmm from', args.load_ubm
ubm_gmm = loadUBM(args.load_ubm)
#####
# Enrollment
# now for each feature-set adapt a gmm
#####
descriptor_files, labels = pc.getFiles(args.inputfolder, args.suffix,
args.labelfile)
if len(descriptor_files) == 0:
print 'no descriptor_files'
sys.exit(1)
elif labels:
num_scribes = len(list(set(labels)))
else:
num_scribes = 'unknown'
num_descr = len(descriptor_files)
print 'number of classes:', num_scribes
print 'number of descriptor_files:', num_descr
print 'adapt traing-features to create individual scribe-gmms (or load saved ones)'
widgets = [progressbar.Percentage(), ' ', progressbar.Bar(), ' ',
progressbar.ETA()]
progress = progressbar.ProgressBar(widgets=widgets,
maxval=len(descriptor_files))
if args.encoding == 'supervector':
identifier = '_sv'
elif 'fisher' in args.encoding:
identifier = '_fv'
else: #vlad
identifier = '_vlad'
def encode(i):
if isinstance(descriptor_files[i], basestring):
base = os.path.basename(os.path.splitext(descriptor_files[i])[0])
else:
base = os.path.basename(os.path.commonprefix(descriptor_files[i]))
gmm_name = base + '_gmm.pkl.gz'
gmm = ubm_gmm
# load encoding
if args.load_scores:
filepath = os.path.join(args.load_scores, base + identifier + '.pkl.gz')
if os.path.exists(filepath):
with gzip.open(filepath, 'rb') as f:
enc = cPickle.load(f)
return enc
# load data and preprocess
features = pc.loadDescriptors( descriptor_files[i],
hellinger=args.hellinger,
min_descs_per_file=args.min_descs,
show_progress= True)
if features is None:
print 'WARNING: features==None ?!'
progress.update(i+1)
return 0.0
# make the actual encoding step
enc = encodeGMM(args.encoding, gmm, features,
normalize=args.normalize,
update=args.update, relevance=args.relevance )
# save encoding
filepath = os.path.join(args.outputfolder, base + identifier + '.pkl.gz')
with gzip.open(filepath, 'w') as f:
cPickle.dump(enc, f, -1)
progress.update(i+1)
if args.no_eval: # save some memory
return None
return enc
progress.start()
if args.parallel:
all_enc = zip( *pc.parmap( encode, range(num_descr), args.nprocs ) )
else:
all_enc = zip( *map( encode, range(num_descr) ) )
progress.finish()
print 'got {} encodings'.format(len(all_enc))
if args.no_eval:
sys.exit(1)
all_enc = np.concatenate(all_enc, axis=0).astype(np.float32)
print 'Evaluation:'
ret_matrix = evaluate.runNN( all_enc , labels, parallel=args.parallel,
nprocs=args.nprocs )
if ret_matrix is not None:
fpath = os.path.join(args.outputfolder, 'dist' + identifier + '.cvs')
np.savetxt(fpath, ret_matrix, delimiter=',')