def compute_fbank_pitch():
parser = get_parser()
args = parser.parse_args()
config = {}
config['sample_rate'] = float(args.sample_rate)
config['upper_frequency_limit'] = float(args.upper_frequency_limit)
config['lower_frequency_limit'] = float(args.lower_frequency_limit)
config['filterbank_channel_count'] = float(args.filterbank_channel_count)
config['window_length'] = args.window_length
config['frame_length'] = args.frame_length
config['thres_autoc'] = args.thres_autoc
config['output_type'] = args.output_type
fbank_pitch = FbankPitch.params(config).instantiate()
with kaldiio.ReadHelper(args.rspecifier,
segments=args.segments) as reader, \
KaldiWriter(args.wspecifier, write_num_frames=args.write_num_frames,
compress=args.compress, compression_method=args.compression_method) as writer:
for utt_id, (sample_rate, array) in reader:
if sample_rate != args.sample_rate:
args.sample_rate = sample_rate
array = array.astype(np.float32)
audio_data = tf.constant(array, dtype=tf.float32)
fbank_pitch_test = fbank_pitch(audio_data, args.sample_rate)
sess = tf.Session()
fbank_pitch_feats = fbank_pitch_test.eval(session=sess)
writer[utt_id] = fbank_pitch_feats
def compute_spectrum():
parser = get_parser()
args = parser.parse_args()
config = {}
config['sample_rate'] = float(args.sample_rate)
config['output_type'] = int(args.output_type)
config['window_length'] = args.window_length
config['frame_length'] = args.frame_length
spectrum = Spectrum.params(config).instantiate()
with kaldiio.ReadHelper(args.rspecifier,
segments=args.segments) as reader, \
KaldiWriter(args.wspecifier, write_num_frames=args.write_num_frames,
compress=args.compress, compression_method=args.compression_method) as writer:
for utt_id, (sample_rate, array) in reader:
if sample_rate != args.sample_rate:
args.sample_rate = sample_rate
array = array.astype(np.float32)
audio_data = tf.constant(array, dtype=tf.float32)
spectrum_test = spectrum(audio_data, args.sample_rate)
sess = tf.compat.v1.Session()
spectrum_feats = spectrum_test.eval(session=sess)
writer[utt_id] = spectrum_feats
def compute_mfcc():
parser = get_parser()
args = parser.parse_args()
config = {}
config['sample_rate'] = int(args.sample_rate)
config['upper_frequency_limit'] = float(args.upper_frequency_limit)
config['lower_frequency_limit'] = float(args.lower_frequency_limit)
config['filterbank_channel_count'] = float(args.filterbank_channel_count)
config['window_length'] = args.window_length
config['frame_length'] = args.frame_length
config['output_type'] = args.output_type
config['window_type'] = args.window_type
config['snip_edges'] = args.snip_edges
config['preeph_coeff'] = args.preeph_coeff
config['remove_dc_offset'] = args.remove_dc_offset
config['is_fbank'] = args.is_fbank
config['cepstral_lifter'] = args.cepstral_lifter
config['coefficient_count'] = args.coefficient_count
mfcc = Mfcc.params(config).instantiate()
with kaldiio.ReadHelper(args.rspecifier,
segments=args.segments) as reader, \
KaldiWriter(args.wspecifier, write_num_frames=args.write_num_frames,
compress=args.compress, compression_method=args.compression_method) as writer:
for utt_id, (sample_rate, array) in reader:
if sample_rate != args.sample_rate:
args.sample_rate = sample_rate
array = array.astype(np.float32)
audio_data = tf.constant(array, dtype=tf.float32)
mfcc_test = tf.squeeze(mfcc(audio_data, args.sample_rate))
sess = tf.Session()
mfcc_feats = mfcc_test.eval(session=sess)
writer[utt_id] = mfcc_feats
def compute_stft():
parser = get_parser()
args = parser.parse_args()
config = {}
config['sample_rate'] = int(args.sample_rate)
config['window_length'] = args.window_length
config['frame_length'] = args.frame_length
stft = Analyfiltbank.params(config).instantiate()
with kaldiio.ReadHelper(args.rspecifier,
segments=args.segments) as reader, \
KaldiWriter(args.wspecifier, write_num_frames=args.write_num_frames,
compress=args.compress, compression_method=args.compression_method) as writer:
for utt_id, (sample_rate, array) in reader:
if sample_rate != args.sample_rate:
args.sample_rate = sample_rate
array = array.astype(np.float32)
audio_data = tf.constant(array, dtype=tf.float32)
power_spectrum, phase_spectrum = stft(audio_data, args.sample_rate)
sess = tf.Session()
if args.output_type == 1:
out_feats = power_spectrum.eval(session=sess)
else:
out_feats = phase_spectrum.eval(session=sess)
writer[utt_id] = out_feats
def apply_cmvn():
args = get_parser().parse_args()
if ':' in args.stats_rspecifier_or_rxfilename:
is_rspcifier = True
stats_filetype = 'ark'
stats_dict = dict(KaldiReader(args.stats_rspecifier_or_rxfilename))
else:
is_rspcifier = False
stats_filetype = 'mat'
stats = kaldiio.load_mat(args.stats_rspecifier_or_rxfilename)
stats_dict = {None: stats}
config = {}
config['norm_means'] = args.norm_means
config['norm_vars'] = args.norm_vars
config['utt2spk'] = args.utt2spk
config['spk2utt'] = args.spk2utt
config['reverse'] = args.reverse
config['std_floor'] = args.std_floor
config['filetype'] = stats_filetype
cmvn = CMVN.params(config).instantiate()
cmvn.call(stats_dict)
with KaldiWriter(args.wspecifier, write_num_frames=args.write_num_frames,
compress=args.compress, compression_method=args.compression_method) as writer, \
kaldiio.ReadHelper(args.rspecifier) as reader:
for utt, mat in reader:
mat_new = cmvn.apply_cmvn(mat, utt)
writer[utt] = mat_new
def compute_pitch():
parser = get_parser()
args = parser.parse_args()
config = {}
config['sample_rate'] = int(args.sample_rate)
config['window_length'] = args.window_length
config['frame_length'] = args.frame_length
config['snip_edges'] = args.snip_edges
config['preemph_coeff'] = args.preemph_coeff
config['min_f0'] = args.min_f0
config['max_f0'] = args.max_f0
config['soft_min_f0'] = args.soft_min_f0
config['penalty_factor'] = args.penalty_factor
config['lowpass_cutoff'] = args.lowpass_cutoff
config['resample_freq'] = args.resample_freq
config['delta_pitch'] = args.delta_pitch
config['nccf_ballast'] = args.nccf_ballast
config['lowpass_filter_width'] = args.lowpass_filter_width
config['upsample_filter_width'] = args.upsample_filter_width
config['max_frames_latency'] = args.max_frames_latency
config['frames_per_chunk'] = args.frames_per_chunk
config['simulate_first_pass_online'] = args.simulate_first_pass_online
config['recompute_frame'] = args.recompute_frame
config['nccf_ballast_online'] = args.nccf_ballast_online
pitch = Pitch.params(config).instantiate()
with kaldiio.ReadHelper(args.rspecifier,
segments=args.segments) as reader, \
KaldiWriter(args.wspecifier, write_num_frames=args.write_num_frames,
compress=args.compress, compression_method=args.compression_method) as writer:
for utt_id, (sample_rate, array) in reader:
if sample_rate != args.sample_rate:
args.sample_rate = sample_rate
array = array.astype(np.float32)
audio_data = tf.constant(array, dtype=tf.float32)
pitch_test = tf.squeeze(pitch(audio_data, args.sample_rate))
sess = tf.Session()
pitch_feats = pitch_test.eval(session=sess)
writer[utt_id] = pitch_feats
def compute_cmvn_stats():
"""
e.g. compute_cmvn_stats.py scp:data/train/feats.scp data/train/cmvn.ark # compute global cmvn
"""
args = get_parser().parse_args()
is_wspecifier = ':' in args.wspecifier_or_wxfilename
if is_wspecifier:
if args.spk2utt is not None:
utt2spk_dict = {}
with open(args.spk2utt) as f:
for line in f:
spk, utts = line.rstrip().split(None, 1)
for utt in utts.split():
utt2spk_dict[utt] = spk
def utt2spk(x):
return utt2spk_dict[x]
else:
logging.info('Performing as utterance CMVN mode')
def utt2spk(x):
return x
else:
logging.info('Performing as gloabl CMVN model')
if args.spk2utt is not None:
logging.warning('spk2utt is not used for global CMVN mode')
def utt2spk(x):
return None
# Calculate stats for each speaker
counts = {}
sum_feats = {}
square_sum_feats = {}
idx = 0
for idx, (utt, matrix) in enumerate(KaldiReader(args.rspecifier), 1):
spk = utt2spk(utt)
if spk not in counts:
counts[spk] = 0
feat_shape = matrix.shape[1:]
sum_feats[spk] = np.zeros(feat_shape, dtype=np.float)
square_sum_feats[spk] = np.zeros(feat_shape, dtype=np.float)
counts[spk] += matrix.shape[0]
sum_feats[spk] += matrix.sum(axis=0)
square_sum_feats[spk] += (matrix**2).sum(axis=0)
assert idx > 0, idx
cmvn_stats = {}
for spk in counts:
feat_shape = sum_feats[spk].shape
cmvn_shape = (2, feat_shape[0] + 1) + feat_shape[1:]
_cmvn_stats = np.empty(cmvn_shape, dtype=np.float64)
_cmvn_stats[0, :-1] = sum_feats[spk]
_cmvn_stats[1, :-1] = square_sum_feats[spk]
_cmvn_stats[0, -1] = counts[spk]
_cmvn_stats[1, -1] = 0.
cmvn_stats[spk] = _cmvn_stats
if is_wspecifier:
with KaldiWriter(args.wspecifier_or_wxfilename) as writer:
for spk, mat in cmvn_stats.items():
writer[spk] = mat
else:
matrix = cmvn_stats[None]
kaldiio.save_mat(args.wspecifier_or_wxfilename, matrix)