def _convert(self, audio):
#import speech_recognition as sr
with audio.get_filename() as filename:
with sr.AudioFile(filename) as source:
clip = self.recognizer.record(source)
_json = self._query_api(clip)
if 'results' in _json:
results = _json['results']
else:
raise Exception(
'received invalid results from API: {0}'.format(str(_json)))
elements = []
for result in results:
if result['final'] is True:
timestamps = result['alternatives'][0]['timestamps']
if self.resolution is 'words':
for entry in timestamps:
text = entry[0]
start = entry[1]
end = entry[2]
elements.append(TextStim(text=text, onset=start,
duration=end-start))
elif self.resolution is 'phrases':
text = result['alternatives'][0]['transcript']
start = timestamps[0][1]
end = timestamps[-1][2]
elements.append(TextStim(text=text, onset=start,
duration=end-start))
return ComplexTextStim(elements=elements)
def _convert(self, stim):
request = self._build_request([stim])
responses = self._query_api(request)
response = responses[0]
if response:
annotations = response[self.response_object]
# Combine the annotations
if self.handle_annotations == 'first':
text = annotations[0]['description']
return TextStim(text=text,
onset=stim.onset,
duration=stim.duration)
elif self.handle_annotations == 'concatenate':
text = ''
for annotation in annotations:
text += annotation['description']
return TextStim(text=text,
onset=stim.onset,
duration=stim.duration)
elif self.handle_annotations == 'list':
texts = []
for annotation in annotations:
texts.append(
TextStim(text=annotation['description'],
onset=stim.onset,
duration=stim.duration))
return texts
else:
return TextStim(text='', onset=stim.onset, duration=stim.duration)
def _convert(self, stims):
request = self._build_request(stims)
responses = self._query_api(request)
texts = []
for response in responses:
if response and self.response_object in response:
annotations = response[self.response_object]
# Combine the annotations
if self.handle_annotations == 'first':
text = annotations[0]['description']
texts.append(TextStim(text=text))
elif self.handle_annotations == 'concatenate':
text = ''
for annotation in annotations:
text = ' '.join([text, annotation['description']])
texts.append(TextStim(text=text))
elif self.handle_annotations == 'list':
for annotation in annotations:
texts.append(TextStim(text=annotation['description']))
elif 'error' in response:
raise Exception(response['error']['message'])
else:
texts.append(TextStim(text=''))
return texts
def _filter(self, stim):
if self.tokenize:
tokens = stim.text.split()
stemmed = ' '.join([self.stemmer.stem(tok) for tok in tokens])
else:
stemmed = self.stemmer.stem(stim.text)
return TextStim(stim.filename, stemmed)
def _filter(self, stim):
pos_map = {
'ADJ': 'a',
'ADJ_SAT': 's',
'ADV': 'r',
'NOUN': 'n',
'VERB': 'v'
}
def pos_wordnet(txt):
pos_tagged = dict(nltk.pos_tag(txt, tagset='universal'))
pos_tagged = {t: pos_map[tag] if tag in pos_map else 'n'
for t, tag in pos_tagged.items()}
return pos_tagged
tokens = [stim.text]
if self.tokenize:
tokens = nltk.word_tokenize(tokens[0])
tokens = [t if self.case_sensitive else t.lower() for t in tokens]
if not isinstance(self.stemmer, stem.WordNetLemmatizer):
stemmed = ' '.join([self.stemmer.stem(t) for t in tokens])
else:
pos_tagged = pos_wordnet(tokens)
stemmed = ' '.join([self.stemmer.lemmatize(t, pos=pos_tagged[t])
for t in tokens])
return TextStim(stim.filename, stemmed, stim.onset, stim.duration,
stim.order, stim.url)
def _convert(self, stim):
verify_dependencies(['pytesseract'])
text = pytesseract.image_to_string(Image.fromarray(stim.data))
return TextStim(text=text,
onset=stim.onset,
duration=stim.duration,
order=stim.order)
def _filter(self, stim):
if self.tokenizer:
tokens = self.tokenizer.tokenize(stim.text)
else:
tokens = word_tokenize(stim.text)
stims = [TextStim(stim.filename, token, order=i)
for i, token in enumerate(tokens)]
return stims
def _filter(self, stim):
if self.tokenizer:
tokens = self.tokenizer.tokenize(stim.text)
else:
tokens = word_tokenize(stim.text)
stims = [TextStim(stim.filename, token, stim.onset, stim.duration)
for token in tokens]
return stims
def _filter(self, stim):
tokens = word_tokenize(stim.text)
tokens = [tok for tok in tokens if tok not in self.tokens]
text = ' '.join(tokens)
return TextStim(stim.filename,
text,
onset=stim.onset,
duration=stim.duration,
order=stim.order)
def _filter(self, stim):
if self.tokenize:
tokens = stim.text.split()
stemmed = ' '.join([self.stemmer.stem(tok) for tok in tokens])
else:
stemmed = self.stemmer.stem(stim.text)
return TextStim(stim.filename,
stemmed,
onset=stim.onset,
duration=stim.duration,
order=stim.order)
def _convert(self, stim):
params = {'language': self.language, 'detectOrientation': False}
response = self._query_api(stim, params)
lines = []
for r in response['regions']:
for l in r['lines']:
lines.append(' '.join([w['text'] for w in l['words']]))
text = '\n'.join(lines)
return TextStim(text=text)
def _convert(self, audio):
verify_dependencies(['sr'])
offset = 0.0 if audio.onset is None else audio.onset
with audio.get_filename() as filename:
with sr.AudioFile(filename) as source:
clip = self.recognizer.record(source)
_json = self._query_api(clip)
if 'results' in _json:
results = _json['results']
else:
raise Exception('received invalid results from API: {0}'.format(
str(_json)))
elements = []
order = 0
for result in results:
if result['final'] is True:
timestamps = result['alternatives'][0]['timestamps']
if self.resolution is 'words':
for entry in timestamps:
text = entry[0]
start = entry[1]
end = entry[2]
elements.append(
TextStim(text=text,
onset=offset + start,
duration=end - start,
order=order))
order += 1
elif self.resolution is 'phrases':
text = result['alternatives'][0]['transcript']
start = timestamps[0][1]
end = timestamps[-1][2]
elements.append(
TextStim(text=text,
onset=offset + start,
duration=end - start,
order=order))
order += 1
return ComplexTextStim(elements=elements, onset=audio.onset)
def _convert(self, audio):
import speech_recognition as sr
with sr.AudioFile(audio.filename) as source:
clip = self.recognizer.record(source)
result = self._query_api(clip)
timestamps = result['results'][0]['alternatives'][0]['timestamps']
elements = []
for i, entry in enumerate(timestamps):
elements.append(
TextStim(text=entry[0],
onset=entry[1],
duration=entry[2] - entry[1]))
return ComplexTextStim(elements=elements)
def _convert(self, audio):
verify_dependencies(['rev_ai'])
msg = "Beginning audio transcription with a timeout of %fs. Even for "\
"small audios, full transcription may take awhile." % self.timeout
logging.warning(msg)
if audio.url:
job = self.client.submit_job_url(audio.url)
else:
with audio.get_filename() as filename:
job = self.client.submit_job_local_file(filename)
operation_start = time.time()
response = self.client.get_job_details(job.id)
while (response.status == rev_ai.JobStatus.IN_PROGRESS) and \
(time.time() - operation_start) < self.timeout:
response = self.client.get_job_details(job.id)
time.sleep(self.request_rate)
if (time.time() - operation_start) >= self.timeout:
msg = "Conversion reached the timeout limit of %fs." % self.timeout
logging.warning(msg)
if response.status == rev_ai.JobStatus.FAILED:
raise Exception('API failed: %s' % response.failure_detail)
result = self.client.get_transcript_object(job.id)
elements = []
order = 0
for m in result.monologues:
for e in m.elements:
if e.type_ == 'text':
start = e.timestamp
end = e.end_timestamp
elements.append(
TextStim(text=e.value,
onset=start,
duration=end - start,
order=order))
order += 1
return ComplexTextStim(elements=elements, onset=audio.onset)
def _convert(self, stim):
request = self._build_request(stim)
response = self._query_api(request)
if 'error' in response:
raise Exception(response['error']['message'])
words = []
if 'results' in response:
for result in response['results']:
transcription = result['alternatives'][0]
for w in transcription['words']:
onset = float(w['startTime'][:-1])
duration = float(w['endTime'][:-1]) - onset
words.append(TextStim(text=w['word'],
onset=onset,
duration=duration))
return ComplexTextStim(elements=words)
def _filter(self, stim):
return TextStim(stim.filename, stim.text.lower())
def _filter(self, stim):
stemmed = self.stemmer.stem(stim.text)
return TextStim(stim.filename, stemmed, stim.onset, stim.duration)
def _convert(self, stim):
import pytesseract
text = pytesseract.image_to_string(Image.fromarray(stim.data))
return TextStim(text=text, onset=stim.onset, duration=stim.duration)
def _convert(self, stim):
verify_dependencies(['pytesseract'])
text = pytesseract.image_to_string(Image.fromarray(stim.data))
return TextStim(text=text)
def _filter(self, stim):
tokens = word_tokenize(stim.text)
tokens = [tok for tok in tokens if tok not in self.tokens]
text = ' '.join(tokens)
return TextStim(stim.filename, text)
def _filter(self, stim):
pattern = '[%s]' % re.escape(string.punctuation)
text = re.sub(pattern, '', stim.text)
return TextStim(stim.filename, text)