find pronunciation variants which all phones are in novo70.

This commit is contained in:
yemaozi88 2019-01-07 11:50:24 +01:00
parent dd9e3d820b
commit d6e005b1cb
6 changed files with 187 additions and 12 deletions

Binary file not shown.

View File

@ -35,6 +35,9 @@
<Compile Include="fa_test.py">
<SubType>Code</SubType>
</Compile>
<Compile Include="forced_alignment_novo.py">
<SubType>Code</SubType>
</Compile>
<Compile Include="htk_vs_kaldi.py">
<SubType>Code</SubType>
</Compile>

View File

@ -199,4 +199,4 @@ def make_fame2ipa_variants(fame):
ipa.append(fame.replace('ɔ̈', 'ɒ'))
ipa.append(fame.replace('ɔ̈:', 'ɒ'))
return ipa
return ipa

View File

@ -55,14 +55,15 @@ phoneset_novo70 = np.unique(phoneset_novo70)
# [iː] iy
# [œː] uh
# [ɛː] eh
david_suggestion = ['ɔː', 'ɪː', 'iː', 'œː', 'ɛː']
# [w] wv in IPA written as ʋ.
david_suggestion = ['ɔː', 'ɪː', 'iː', 'œː', 'ɛː', 'w']
## ======================= convert phones ======================
## ======================= extract words which is written only with novo70 ======================
mapping = convert_xsampa2ipa.load_converter('xsampa', 'ipa', default.ipa_xsampa_converter_dir)
stimmen_transcription_ = pd.ExcelFile(default.stimmen_transcription_xlsx)
df = pd.read_excel(stimmen_transcription_, 'check')
df = pd.read_excel(stimmen_transcription_, 'frequency')
#for xsampa, ipa in zip(df['X-SAMPA'], df['IPA']):
# ipa_converted = convert_xsampa2ipa.xsampa2ipa(mapping, xsampa)
# if not ipa_converted == ipa:
@ -70,11 +71,13 @@ df = pd.read_excel(stimmen_transcription_, 'check')
transcription_ipa = list(df['IPA'])
# transcription mistake?
transcription_ipa = [ipa.replace(';', ':') for ipa in transcription_ipa if not ipa=='pypɪl' and not pd.isnull(ipa)]
transcription_ipa = [ipa.replace(';', 'ː') for ipa in transcription_ipa if not ipa=='pypɪl' and not pd.isnull(ipa)]
transcription_ipa = [ipa.replace('ˑ', '') for ipa in transcription_ipa] # only one case.
not_in_novo70 = []
all_in_novo70 = []
for ipa in transcription_ipa:
ipa = ipa.replace(':', 'ː')
ipa = convert_phone_set.split_ipa(ipa)
not_in_novo70_ = [phone for phone in ipa
@ -83,19 +86,76 @@ for ipa in transcription_ipa:
not_in_novo70_ = [phone.replace(':', '') for phone in not_in_novo70_]
not_in_novo70_ = [phone.replace('ː', '') for phone in not_in_novo70_]
if len(not_in_novo70_) == 0:
all_in_novo70.append(''.join(ipa))
#translation_key.get(phone, phone)
not_in_novo70.extend(not_in_novo70_)
not_in_novo70_list = list(set(not_in_novo70))
## check which phone is used in stimmen but not in novo70
# 'ʀ', 'ʁ',
# 'ɒ', 'ɐ',
# 'o', 'a' (o:, a:?)
# [e] 'nyːver mɑntsjə' (1)
# [ɾ] 'ɪːɾ'(1)
# [ɹ] 'iːjəɹ' (1), 'ɪ:ɹ' (1)
# [ø] 'gʀøtəpi:r'(1), 'grøtəpi:r'(1)
# [æ] 'røːzəʀæt'(2), 'røːzəræt'(1)
# [ʊ] 'ʊ'(1) --> can be ʏ (uh)??
# [χ] --> can be x??
def search_phone_ipa(x, phone_list):
return [phone for phone in phone_list if x in convert_phone_set.split_ipa(phone)]
x_in_item = []
for ipa in phone_list:
ipa_original = ipa
ipa = ipa.replace(':', 'ː')
ipa = convert_phone_set.split_ipa(ipa)
if x in ipa and not x+':' in ipa:
x_in_item.append(ipa_original)
return x_in_item
#search_phone_ipa('ø', transcription_ipa)
# 'ɐ', 'ɒ', 'w', 'æ', 'ʀ', 'ʁ',
# 'œː', 'ɾ',
# 'o', 'a'
# [e] 'nyːver mɑntsjə' (1)
# [ɹ] 'iːjəɹ' (2)
search_phone_ipa('ˑ', transcription_ipa)
df = pd.read_excel(stimmen_transcription_, 'original')
ipas = []
famehtks = []
for xsampa in df['Self Xsampa']:
if not isinstance(xsampa, float): # 'NaN'
# typo?
xsampa = xsampa.replace('r2:z@rA:\\t', 'r2:z@rA:t')
xsampa = xsampa.replace(';', ':')
ipa = convert_xsampa2ipa.xsampa2ipa(mapping, xsampa)
ipa = ipa.replace('ː', ':')
ipa = ipa.replace(' ', '')
ipas.append(ipa)
else:
ipas.append('')
# extract interesting cols.
df = pd.DataFrame({'filename': df['Filename'],
'word': df['Word'],
'xsampa': df['Self Xsampa'],
'ipa': pd.Series(ipas)})
# find options which all phones are in novo70.
#word_list = list(set(df['word']))
#word_list = [word for word in word_list if not pd.isnull(word)]
#word = word_list[1]
## pronunciation variants of 'word'
#df_ = df[df['word'] == word]['xsampa']
##pronunciation_variant = list(set(df_))
cols = ['word', 'ipa', 'frequency']
df_samples = pd.DataFrame(index=[], columns=cols)
for ipa in all_in_novo70:
ipa = ipa.replace('ː', ':')
samples = df[df['ipa'] == ipa]
word = list(set(samples['word']))[0]
samples_Series = pd.Series([word, ipa, len(samples)], index=df_samples.columns)
df_samples = df_samples.append(samples_Series, ignore_index=True)

View File

@ -0,0 +1,112 @@
#
# forced alignment using novo-api.
#
# *** IMPORTANT ***
# This file should be treated as confidencial.
# This file should not be copied or uploaded to public sites.
#
# NOTES:
# The usage of novo api: https://bitbucket.org/novolanguage/python-novo-api
# I couldn't make it work as I described in the mail to Martijn Bartelds on
# 2018/12/03.
# As per the advice from him, I modified testgrammer.py and made it a function.
#
# In order to run on Python 3.6, the following points are changed in novo-api.
# (1) backend/__init__.py
# - #import session
# from . import session
# (2) backend/session.py
# - #except Exception, e:
# except Exception as e:
# - #print self.last_message
# print(self.last_message)
# (3) asr/segment/praat.py
# - def print_tier(output, title, begin, end, segs, (format, formatter))
# def print_tier(output, title, begin, end, segs, format, formatter):
# (4) asr/spraaklab/__init.py
# - #import session
# from . import session
# (5) asr/spraaklab/schema.py
# - #print data, "validated not OK", e.message
# print("{0} validated not OK {1}".format(data, e.message))
# - #print data, "validated OK"
# print("{} validated OK".format(data))
# - #if isinstance(object, basestring):
# if isinstance(object, str)
#
# Aki Kunikoshi
# 428968@gmail.com
#
import argparse
import json
from novoapi.backend import session
# username / password cannot be passed as artuments...
p = argparse.ArgumentParser()
#p.add_argument("--user", default=None)
#p.add_argument("--password", default=None)
p.add_argument("--user", default='martijn.wieling')
p.add_argument("--password", default='fa0Thaic')
args = p.parse_args()
wav_file = 'c:\\OneDrive\\WSL\\test\\onetwothree.wav'
rec = session.Recognizer(grammar_version="1.0", lang="nl", snodeid=101, user=args.user, password=args.password, keepopen=True) # , modeldir=modeldir)
grammar = {
"type": "confusion_network",
"version": "1.0",
"data": {
"kind": "sequence",
"elements": [{
"kind": "word",
"pronunciation": [{
"phones": ["wv",
"a1",
"n"],
"id": 0
},
{
"phones": ["wv",
"uh1",
"n"],
"id": 1
}],
"label": "one"
},
{
"kind": "word",
"pronunciation": [{
"phones": ["t",
"uw1"],
"id": 0
}],
"label": "two"
},
{
"kind": "word",
"pronunciation": [{
"phones": ["t",
"r",
"iy1"],
"id": 0
},
{
"phones": ["s",
"r",
"iy1"],
"id": 1
}],
"label": "three"
}]
},
"return_objects": ["grammar"],
"phoneset": "novo70"
}
res = rec.setgrammar(grammar)
#print "Set grammar result", res
#res = rec.recognize_wav("test/onetwothree.wav")
res = rec.recognize_wav(wav_file)
#print "Recognition result:", json.dumps(res.export(), indent=4)