def main(*argv):
argv = argv if argv else sys.argv[1:]
# Options for python
description = 'estimate joint feature of source and target speakers'
parser = argparse.ArgumentParser(description=description)
parser.add_argument('speaker', type=str, help='Input speaker label')
parser.add_argument('list_file',
type=str,
help='List file of the input speaker')
parser.add_argument('pair_dir',
type=str,
help='Statistics directory of the speaker')
args = parser.parse_args(argv)
# open h5 files
h5_dir = os.path.join(args.pair_dir, 'h5')
statspath = os.path.join(args.pair_dir, 'stats', args.speaker + '.h5')
h5 = HDF5(statspath, mode='a')
# estimate and save F0 statistics
f0stats = F0statistics()
f0s = read_feats(args.list_file, h5_dir, ext='f0')
f0stats = f0stats.estimate(f0s)
h5.save(f0stats, ext='f0stats')
print("f0stats save into " + statspath)
# estimate and save GV of orginal and target speakers
gv = GV()
mceps = read_feats(args.list_file, h5_dir, ext='mcep')
gvstats = gv.estimate(mceps)
h5.save(gvstats, ext='gv')
print("gvstats save into " + statspath)
h5.close()
def main(*argv):
argv = argv if argv else sys.argv[1:]
# Options for python
description = 'estimate joint feature of source and target speakers'
parser = argparse.ArgumentParser(description=description)
parser.add_argument('speaker', type=str,
help='Input speaker label')
parser.add_argument('list_file', type=str,
help='List file of the input speaker')
parser.add_argument('pair_dir', type=str,
help='Statistics directory of the speaker')
args = parser.parse_args(argv)
# open h5 files
h5_dir = os.path.join(args.pair_dir, 'h5')
statspath = os.path.join(args.pair_dir, 'stats', args.speaker + '.h5')
h5 = HDF5(statspath, mode='w')
# estimate and save F0 statistics
f0stats = F0statistics()
f0s = read_feats(args.list_file, h5_dir, ext='f0')
f0stats = f0stats.estimate(f0s)
h5.save(f0stats, ext='f0stats')
print("f0stats save into " + statspath)
# estimate and save GV of orginal and target speakers
gv = GV()
mceps = read_feats(args.list_file, h5_dir, ext='mcep')
gvstats = gv.estimate(mceps)
h5.save(gvstats, ext='gv')
print("gvstats save into " + statspath)
h5.close()
def main(*argv):
argv = argv if argv else sys.argv[1:]
# Options for python
description = 'Train GMM and converted GV statistics'
parser = argparse.ArgumentParser(description=description)
parser.add_argument('org_list_file',
type=str,
help='List file of original speaker')
parser.add_argument('pair_yml',
type=str,
help='Yml file of the speaker pair')
parser.add_argument('pair_dir',
type=str,
help='Directory path of h5 files')
args = parser.parse_args(argv)
# read pair-dependent yml file
pconf = PairYML(args.pair_yml)
# read joint feature vector
jntf = os.path.join(args.pair_dir, 'jnt',
'it' + str(pconf.jnt_n_iter + 1) + '_jnt.h5')
jnth5 = HDF5(jntf, mode='r')
jnt = jnth5.read(ext='jnt')
# train GMM for mcep using joint feature vector
gmm = GMMTrainer(n_mix=pconf.GMM_mcep_n_mix,
n_iter=pconf.GMM_mcep_n_iter,
covtype=pconf.GMM_mcep_covtype)
gmm.train(jnt)
# save GMM
gmm_dir = os.path.join(args.pair_dir, 'model')
if not os.path.exists(gmm_dir):
os.makedirs(gmm_dir)
gmmpath = os.path.join(gmm_dir, 'GMM.pkl')
joblib.dump(gmm.param, gmmpath)
print("Conversion model save into " + gmmpath)
# calculate GV statistics of converted feature
h5_dir = os.path.join(args.pair_dir, 'h5')
org_mceps = read_feats(args.org_list_file, h5_dir, ext='mcep')
cv_mceps = feature_conversion(pconf, org_mceps, gmm, gmmmode=None)
diffcv_mceps = feature_conversion(pconf, org_mceps, gmm, gmmmode='diff')
gv = GV()
cvgvstats = gv.estimate(cv_mceps)
diffcvgvstats = gv.estimate(diffcv_mceps)
# open h5 files
statspath = os.path.join(gmm_dir, 'cvgv.h5')
cvgvh5 = HDF5(statspath, mode='w')
cvgvh5.save(cvgvstats, ext='cvgv')
cvgvh5.save(diffcvgvstats, ext='diffcvgv')
print("Converted gvstats save into " + statspath)
def test_GVstatistics(self):
gvstats = GV()
datalist = []
for i in range(1, 4):
datalist.append(np.random.rand(100 * i).reshape(100 * i // 2, 2))
gv = gvstats.estimate(datalist)
data = np.random.rand(100 * 5).reshape(100 * 5 // 2, 2)
odata = gvstats.postfilter(data, gv, startdim=0)
assert data.shape[0] == odata.shape[0]
odata = gvstats.postfilter(data, gv, alpha=0.0, startdim=0)
assert np.all(data == odata)
def test_GVstatistics(self):
gvstats = GV()
datalist = []
for i in range(1, 4):
datalist.append(np.random.rand(100 * i).reshape(100 * i // 2, 2))
gv = gvstats.estimate(datalist)
data = np.random.rand(100 * 5).reshape(100 * 5 // 2, 2)
odata = gvstats.postfilter(data, gv, startdim=0)
assert data.shape[0] == odata.shape[0]
odata = gvstats.postfilter(data, gv, alpha=0.0, startdim=0)
assert np.all(data == odata)
def main(*argv):
argv = argv if argv else sys.argv[1:]
# Options for python
description = 'estimate joint feature of source and target speakers'
parser = argparse.ArgumentParser(description=description)
parser.add_argument('-gmmmode',
'--gmmmode',
type=str,
default=None,
help='mode of the GMM [None, diff, or intra]')
parser.add_argument('org', type=str, help='Original speaker')
parser.add_argument('tar', type=str, help='Target speaker')
parser.add_argument('org_yml',
type=str,
help='Yml file of the original speaker')
parser.add_argument('pair_yml',
type=str,
help='Yml file of the speaker pair')
parser.add_argument('eval_list_file',
type=str,
help='List file for evaluation')
parser.add_argument('wav_dir',
type=str,
help='Directory path of source spekaer')
parser.add_argument('pair_dir',
type=str,
help='Directory path of pair directory')
args = parser.parse_args(argv)
# read parameters from speaker yml
sconf = SpeakerYML(args.org_yml)
pconf = PairYML(args.pair_yml)
# read GMM for mcep
mcepgmmpath = os.path.join(args.pair_dir, 'model/GMM_mcep.pkl')
mcepgmm = GMMConvertor(
n_mix=pconf.GMM_mcep_n_mix,
covtype=pconf.GMM_mcep_covtype,
gmmmode=args.gmmmode,
)
param = joblib.load(mcepgmmpath)
mcepgmm.open_from_param(param)
print("GMM for mcep conversion mode: {}".format(args.gmmmode))
# read F0 statistics
stats_dir = os.path.join(args.pair_dir, 'stats')
orgstatspath = os.path.join(stats_dir, args.org + '.h5')
orgstats_h5 = HDF5(orgstatspath, mode='r')
orgf0stats = orgstats_h5.read(ext='f0stats')
orgstats_h5.close()
# read F0 and GV statistics for target
tarstatspath = os.path.join(stats_dir, args.tar + '.h5')
tarstats_h5 = HDF5(tarstatspath, mode='r')
tarf0stats = tarstats_h5.read(ext='f0stats')
targvstats = tarstats_h5.read(ext='gv')
tarstats_h5.close()
# read GV statistics for converted mcep
cvgvstatspath = os.path.join(args.pair_dir, 'model', 'cvgv.h5')
cvgvstats_h5 = HDF5(cvgvstatspath, mode='r')
cvgvstats = cvgvstats_h5.read(ext='cvgv')
diffcvgvstats = cvgvstats_h5.read(ext='diffcvgv')
cvgvstats_h5.close()
mcepgv = GV()
f0stats = F0statistics()
# constract FeatureExtractor class
feat = FeatureExtractor(analyzer=sconf.analyzer,
fs=sconf.wav_fs,
fftl=sconf.wav_fftl,
shiftms=sconf.wav_shiftms,
minf0=sconf.f0_minf0,
maxf0=sconf.f0_maxf0)
# constract Synthesizer class
synthesizer = Synthesizer(fs=sconf.wav_fs,
fftl=sconf.wav_fftl,
shiftms=sconf.wav_shiftms)
# test directory
test_dir = os.path.join(args.pair_dir, 'test')
os.makedirs(os.path.join(test_dir, args.org), exist_ok=True)
# conversion in each evaluation file
with open(args.eval_list_file, 'r') as fp:
for line in fp:
# open wav file
f = line.rstrip()
wavf = os.path.join(args.wav_dir, f + '.wav')
fs, x = wavfile.read(wavf)
x = x.astype(np.float)
x = low_cut_filter(x, fs, cutoff=70)
assert fs == sconf.wav_fs
# analyze F0, mcep, and ap
f0, spc, ap = feat.analyze(x)
mcep = feat.mcep(dim=sconf.mcep_dim, alpha=sconf.mcep_alpha)
mcep_0th = mcep[:, 0]
# convert F0
cvf0 = f0stats.convert(f0, orgf0stats, tarf0stats)
# convert mcep
cvmcep_wopow = mcepgmm.convert(static_delta(mcep[:, 1:]),
cvtype=pconf.GMM_mcep_cvtype)
cvmcep = np.c_[mcep_0th, cvmcep_wopow]
# synthesis VC w/ GV
if args.gmmmode is None:
cvmcep_wGV = mcepgv.postfilter(cvmcep,
targvstats,
cvgvstats=cvgvstats,
alpha=pconf.GV_morph_coeff,
startdim=1)
wav = synthesizer.synthesis(
cvf0,
cvmcep_wGV,
ap,
rmcep=mcep,
alpha=sconf.mcep_alpha,
)
wavpath = os.path.join(test_dir, f + '_VC.wav')
# synthesis DIFFVC w/ GV
if args.gmmmode == 'diff':
cvmcep[:, 0] = 0.0
cvmcep_wGV = mcepgv.postfilter(mcep + cvmcep,
targvstats,
cvgvstats=diffcvgvstats,
alpha=pconf.GV_morph_coeff,
startdim=1) - mcep
wav = synthesizer.synthesis_diff(
x,
cvmcep_wGV,
rmcep=mcep,
alpha=sconf.mcep_alpha,
)
wavpath = os.path.join(test_dir, f + '_DIFFVC.wav')
# write waveform
if not os.path.exists(os.path.join(test_dir, f)):
os.makedirs(os.path.join(test_dir, f))
wav = np.clip(wav, -32768, 32767)
wavfile.write(wavpath, fs, wav.astype(np.int16))
print(wavpath)
def main(*argv):
argv = argv if argv else sys.argv[1:]
# Options for python
description = 'Train GMM and converted GV statistics'
parser = argparse.ArgumentParser(description=description)
parser.add_argument('org_list_file', type=str,
help='List file of original speaker')
parser.add_argument('pair_yml', type=str,
help='Yml file of the speaker pair')
parser.add_argument('pair_dir', type=str,
help='Directory path of h5 files')
args = parser.parse_args(argv)
# read pair-dependent yml file
pconf = PairYML(args.pair_yml)
# read joint feature vector
jntf = os.path.join(args.pair_dir, 'jnt',
'it' + str(pconf.jnt_n_iter + 1) + '_jnt.h5')
jnth5 = HDF5(jntf, mode='r')
jnt = jnth5.read(ext='mcep')
jnt_codeap = jnth5.read(ext='codeap')
# train GMM for mcep using joint feature vector
gmm = GMMTrainer(n_mix=pconf.GMM_mcep_n_mix,
n_iter=pconf.GMM_mcep_n_iter,
covtype=pconf.GMM_mcep_covtype)
gmm.train(jnt)
# train GMM for codeap using joint feature vector
gmm_codeap = GMMTrainer(n_mix=pconf.GMM_codeap_n_mix,
n_iter=pconf.GMM_codeap_n_iter,
covtype=pconf.GMM_codeap_covtype)
gmm_codeap.train(jnt_codeap)
# save GMM
gmm_dir = os.path.join(args.pair_dir, 'model')
if not os.path.exists(gmm_dir):
os.makedirs(gmm_dir)
gmmpath = os.path.join(gmm_dir, 'GMM_mcep.pkl')
joblib.dump(gmm.param, gmmpath)
print("Conversion model for mcep save into " + gmmpath)
gmmpath_codeap = os.path.join(gmm_dir, 'GMM_codeap.pkl')
joblib.dump(gmm_codeap.param, gmmpath_codeap)
print("Conversion model for codeap save into " + gmmpath_codeap)
# calculate GV statistics of converted feature
h5_dir = os.path.join(args.pair_dir, 'h5')
org_mceps = read_feats(args.org_list_file, h5_dir, ext='mcep')
cv_mceps = feature_conversion(pconf, org_mceps, gmm, gmmmode=None)
diffcv_mceps = feature_conversion(pconf, org_mceps, gmm, gmmmode='diff')
gv = GV()
cvgvstats = gv.estimate(cv_mceps)
diffcvgvstats = gv.estimate(diffcv_mceps)
# open h5 files
statspath = os.path.join(gmm_dir, 'cvgv.h5')
cvgvh5 = HDF5(statspath, mode='w')
cvgvh5.save(cvgvstats, ext='cvgv')
cvgvh5.save(diffcvgvstats, ext='diffcvgv')
print("Converted gvstats save into " + statspath)