def __init__(self, args):
validation_data_name = args.data_root + '/' + args.test_data
trials_filename = args.data_root + '/' + args.test_trials
# Load validation trials
tst_trials = []
tst_enrolls = []
tst_tests = []
for line in open(trials_filename, 'r').readlines():
row = line.split()
tst_enrolls.append(row[0])
tst_tests.append(row[1])
tst_trials.append(np.where(row[2] == 'target', 1, 0))
tst_trials = np.array(tst_trials)
wavlist, utt_label, spk_label = kd.read_data_list(validation_data_name,
utt2spk=True,
utt2lang=False)
tst_dict = dict()
tst_list = []
for value, key in enumerate(utt_label):
tst_dict[key] = value
tst_enrolls_idx = np.array([], dtype=int)
tst_tests_idx = np.array([], dtype=int)
for index in range(len(tst_enrolls)):
tst_enrolls_idx = np.append(tst_enrolls_idx,
int(tst_dict[tst_enrolls[index]]))
tst_tests_idx = np.append(tst_tests_idx,
int(tst_dict[tst_tests[index]]))
self.wavlist = wavlist
self.tst_enrolls_idx = tst_enrolls_idx
self.tst_tests_idx = tst_tests_idx
self.data = validation_data_name
self.tst_trials = tst_trials
self.args = args
help="source data")
parser.add_argument("--split",
type=str,
default="result_test_sorted.csv",
help="target data")
parser.add_argument("--utt2spk", action='store_true', help="for utt2spk file")
parser.add_argument("--utt2lang",
action='store_true',
help="for utt2lang file")
args = parser.parse_known_args()[0]
SOURCE_FOLDER = args.source
TOTAL_SPLIT = int(args.split)
if args.utt2spk:
wavlist, utt_label, spk_label = kd.read_data_list(SOURCE_FOLDER,
utt2spk=True)
if args.utt2lang:
wavlist, utt_label, lang_label = kd.read_data_list(SOURCE_FOLDER,
utt2lang=True)
if args.utt2spk:
kd.split_data(SOURCE_FOLDER,
wavlist,
utt_label,
spk_label=spk_label,
total_split=TOTAL_SPLIT)
if args.utt2lang:
kd.split_data(SOURCE_FOLDER,
wavlist,
utt_label,
parser.add_argument("--utt2lang",
action='store_true',
help="for utt2lang file")
args = parser.parse_known_args()[0]
wavlist = []
utt_label = []
spk_label = []
lang_label = []
for name in [args.source1, args.source2]:
print name
if args.utt2spk:
if args.utt2lang:
wav, utt, spk, lang = kd.read_data_list(name,
utt2spk=True,
utt2lang=True)
lang_label.extend(lang)
else:
wav, utt, spk = kd.read_data_list(name, utt2spk=True)
spk_label.extend(spk)
elif args.utt2lang:
wav, utt, lang = kd.read_data_list(name, utt2lang=True)
lang_label.extend(lang)
wavlist.extend(wav)
utt_label.extend(utt)
# if args.utt2lang:
# wavlist,utt_label,lang_label = kd.read_data_list(args.source1,utt2lang=True)
#
type=str,
default="data/test_segments/utt2lang_sorted",
help="source data")
parser.add_argument("--target",
type=str,
default="result_test_sorted.csv",
help="target data")
parser.add_argument("--utt2spk", action='store_true', help="for utt2spk file")
parser.add_argument("--utt2lang",
action='store_true',
help="for utt2lang file")
args = parser.parse_known_args()[0]
SOURCE_FOLDER = args.source
TARGET_FOLDER = args.target
if not os.path.exists(TARGET_FOLDER):
os.mkdir(TARGET_FOLDER)
wavlist, utt_label, spk_label = kd.read_data_list(SOURCE_FOLDER,
utt2spk=args.utt2spk,
utt2lang=args.utt2lang)
idx = range(len(wavlist))
np.random.shuffle(idx)
wavlist = wavlist[idx]
utt_label = utt_label[idx]
spk_label = spk_label[idx]
kd.write_data(TARGET_FOLDER, wavlist, utt_label, spk_label)
import sys sys.path.insert(0,'scripts/') import kaldi_data as kd BASE_FOLDER = sys.argv[1] TOTAL_SPLIT = int(sys.argv[2]) wavlist,utt_label,lang_label = kd.read_data_list(BASE_FOLDER,utt2lang=True) kd.split_data(BASE_FOLDER,wavlist,utt_label,lang_label=lang_label,total_split=TOTAL_SPLIT)
WIN_LENGTH = int(args.win_len)
# FIXED_LEN = int(args.fixed_len) #298
SOFTMAX_NUM = args.softmax_num
RESUME_STARTPOINT = args.resume_startpoint
NN_MODEL = args.model_name
EMBEDDING_LAYER = args.embedding_layer
if VAD =='False':
VAD = False
if CMVN == 'False':
CMVN = False
is_batchnorm = True
if not args.segments_format:
if int(TOTAL_SPLIT)==1:
wavlist,utt_label,spk_label = kd.read_data_list(DATA_FOLDER, utt2spk=True)
else:
wavlist,utt_label,spk_label = kd.read_data_list(DATA_FOLDER+'/split'+TOTAL_SPLIT+'/'+CURRRENT_SPLIT, utt2spk=True)
feat, _, utt_shape, tffilename = ft.feat_extract(wavlist,FEAT_TYPE,N_FFT,HOP,VAD,CMVN,EXCLUDE_SHORT)
else:
if int(TOTAL_SPLIT)==1:
wavlist,utt_label,seg_wavlist,seg_segid,seg_uttid,seg_windows = kd.read_data_list(DATA_FOLDER, utt2spk=False,segments=True)
else:
wavlist,utt_label,seg_wavlist,seg_segid,seg_uttid,seg_windows = kd.read_data_list(DATA_FOLDER+'/split'+TOTAL_SPLIT+'/'+CURRRENT_SPLIT, utt2spk=False,segments=True)
feat, _, utt_shape, tffilename = ft.feat_extract(seg_wavlist,FEAT_TYPE,N_FFT,HOP,VAD,CMVN,EXCLUDE_SHORT,seg_windows=seg_windows)
SAVER_FOLDERNAME = 'saver/'+NN_MODEL+'_'+tffilename
nn_model = __import__(NN_MODEL)
x = tf.placeholder(tf.float32, [None,None,FEAT_DIM])
import numpy as np
import os, sys
sys.path.insert(0, 'scripts/')
import kaldi_data as kd
import argparse
parser = argparse.ArgumentParser(description="Shuttling data", add_help=True)
parser.add_argument("--source",
type=str,
default="data/voxceleb1_dev",
help="source data")
parser.add_argument("--target",
type=str,
default="data/voxceleb1_dev_1utt",
help="target data")
args = parser.parse_known_args()[0]
if not os.path.exists(args.target):
os.mkdir(args.target)
wavlist, utt_label, spk_label = kd.read_data_list(args.source,
utt2spk=True,
utt2lang=False)
_, idx = np.unique(spk_label, return_index=True)
wavlist = wavlist[idx]
utt_label = utt_label[idx]
spk_label = spk_label[idx]
kd.write_data(args.target, wavlist, utt_label, spk_label)