@ -8,6 +8,7 @@ import json
import random import random
import soundfile import soundfile
import numpy as np import numpy as np
import itertools
import os import os
RANDOM_SEED = 0 RANDOM_SEED = 0
@ -62,6 +63,7 @@ class DataGenerator(object):
self . __stride_ms__ = stride_ms
self . __stride_ms__ = stride_ms
self . __window_ms__ = window_ms
self . __window_ms__ = window_ms
self . __max_frequency__ = max_frequency
self . __max_frequency__ = max_frequency
self . __epoc__ = 0
self . __random__ = random . Random ( RANDOM_SEED )
self . __random__ = random . Random ( RANDOM_SEED )
# load vocabulary (dictionary)
# load vocabulary (dictionary)
self . __vocab_dict__ , self . __vocab_list__ = \
self . __vocab_dict__ , self . __vocab_list__ = \
@ -245,10 +247,42 @@ class DataGenerator(object):
new_batch . append ( ( padded_audio , text ) )
new_batch . append ( ( padded_audio , text ) )
return new_batch
return new_batch
def instance_reader_creator ( self ,
def __batch_shuffle__ ( self , manifest , batch_size ) :
manifest_path ,
"""
sort_by_duration = True ,
The instances have different lengths and they cannot be
shuffle = False ) :
combined into a single matrix multiplication . It usually
sorts the training examples by length and combines only
similarly - sized instances into minibatches , pads with
silence when necessary so that all instances in a batch
have the same length . This batch shuffle fuction is used
to make similarly - sized instances into minibatches and
make a batch - wise shuffle .
1. Sort the audio clips by duration .
2. Generate a random number ` k ` , k in [ 0 , batch_size ) .
3. Randomly remove ` k ` instances in order to make different mini - batches ,
then make minibatches and each minibatch size is batch_size .
4. Shuffle the minibatches .
: param manifest : manifest file .
: type manifest : list
: param batch_size : Batch size . This size is also used for generate
a random number for batch shuffle .
: type batch_size : int
: return : batch shuffled mainifest .
: rtype : list
"""
manifest . sort ( key = lambda x : x [ " duration " ] )
shift_len = self . __random__ . randint ( 0 , batch_size - 1 )
batch_manifest = zip ( * [ iter ( manifest [ shift_len : ] ) ] * batch_size )
self . __random__ . shuffle ( batch_manifest )
batch_manifest = list ( sum ( batch_manifest , ( ) ) )
res_len = len ( manifest ) - shift_len - len ( batch_manifest )
batch_manifest . extend ( manifest [ - res_len : ] )
batch_manifest . extend ( manifest [ 0 : shift_len ] )
return batch_manifest
def instance_reader_creator ( self , manifest ) :
"""
"""
Instance reader creator for audio data . Creat a callable function to
Instance reader creator for audio data . Creat a callable function to
produce instances of data .
produce instances of data .
@ -256,32 +290,13 @@ class DataGenerator(object):
Instance : a tuple of a numpy ndarray of audio spectrogram and a list of
Instance : a tuple of a numpy ndarray of audio spectrogram and a list of
tokenized and indexed transcription text .
tokenized and indexed transcription text .
: param manifest_path : Filepath of manifest for audio clip files .
: param manifest : Filepath of manifest for audio clip files .
: type manifest_path : basestring
: type manifest : basestring
: param sort_by_duration : Sort the audio clips by duration if set True
( for SortaGrad ) .
: type sort_by_duration : bool
: param shuffle : Shuffle the audio clips if set True .
: type shuffle : bool
: return : Data reader function .
: return : Data reader function .
: rtype : callable
: rtype : callable
"""
"""
if sort_by_duration and shuffle :
sort_by_duration = False
logger . warn ( " When shuffle set to true, "
" sort_by_duration is forced to set False. " )
def reader ( ) :
def reader ( ) :
# read manifest
manifest = self . __read_manifest__ (
manifest_path = manifest_path ,
max_duration = self . __max_duration__ ,
min_duration = self . __min_duration__ )
# sort (by duration) or shuffle manifest
if sort_by_duration :
manifest . sort ( key = lambda x : x [ " duration " ] )
if shuffle :
self . __random__ . shuffle ( manifest )
# extract spectrogram feature
# extract spectrogram feature
for instance in manifest :
for instance in manifest :
spectrogram = self . __audio_featurize__ (
spectrogram = self . __audio_featurize__ (
@ -296,8 +311,8 @@ class DataGenerator(object):
batch_size ,
batch_size ,
padding_to = - 1 ,
padding_to = - 1 ,
flatten = False ,
flatten = False ,
sort _by_duration= Tru e,
sort agrad= Fals e,
= False ) :
batch_ shuffle= False ) :
"""
"""
Batch data reader creator for audio data . Creat a callable function to
Batch data reader creator for audio data . Creat a callable function to
produce batches of data .
produce batches of data .
@ -317,20 +332,32 @@ class DataGenerator(object):
: param flatten : If set True , audio data will be flatten to be a 1 - dim
: param flatten : If set True , audio data will be flatten to be a 1 - dim
ndarray . Otherwise , 2 - dim ndarray . Default is False .
ndarray . Otherwise , 2 - dim ndarray . Default is False .
: type flatten : bool
: type flatten : bool
: param sort_by_duration : Sort the audio clips by duration if set True
: param sortagrad : Sort the audio clips by duration in the first epoc
( for SortaGrad ) .
if set True .
: type sort_by_duration : bool
: type sortagrad : bool
: param shuffle : Shuffle the audio clips if set True .
: param batch_shuffle : Shuffle the audio clips if set True . It is
: type shuffle : bool
not a thorough instance - wise shuffle , but a
specific batch - wise shuffle . For more details ,
please see ` __batch_shuffle__ ` function .
: type batch_shuffle : bool
: return : Batch reader function , producing batches of data when called .
: return : Batch reader function , producing batches of data when called .
: rtype : callable
: rtype : callable
"""
"""
def batch_reader ( ) :
def batch_reader ( ) :
instance_reader = self . instance_reader_creator (
# read manifest
manifest = self . __read_manifest__ (
manifest_path = manifest_path ,
manifest_path = manifest_path ,
sort_by_duration = sort_by_duration ,
max_duration = self . __max_duration__ ,
shuffle = shuffle )
min_duration = self . __min_duration__ )
# sort (by duration) or shuffle manifest
if self . __epoc__ == 0 and sortagrad :
manifest . sort ( key = lambda x : x [ " duration " ] )
elif batch_shuffle :
manifest = self . __batch_shuffle__ ( manifest , batch_size )
instance_reader = self . instance_reader_creator ( manifest )
batch = [ ]
batch = [ ]
for instance in instance_reader ( ) :
for instance in instance_reader ( ) :
batch . append ( instance )
batch . append ( instance )
@ -339,6 +366,7 @@ class DataGenerator(object):
batch = [ ]
batch = [ ]
if len ( batch ) > 0 :
if len ( batch ) > 0 :
yield self . __padding_batch__ ( batch , padding_to , flatten )
yield self . __padding_batch__ ( batch , padding_to , flatten )
self . __epoc__ + = 1
return batch_reader
return batch_reader
yangyaming
8 years ago