Is there any way to get timestamps for sentences rather than words in text transcripts received from Google Speech API? I am currently implementing the following code.
#this python file deals with providing word offsets for seek operations
def transcribe_gcs(gcs_uri):
"""Asynchronously transcribes the audio file specified by the gcs_uri."""
from google.cloud import speech
from google.cloud.speech import enums
from google.cloud.speech import types
client = speech.SpeechClient()
audio = types.RecognitionAudio(uri=gcs_uri)
config = types.RecognitionConfig(
encoding=enums.RecognitionConfig.AudioEncoding.FLAC,
sample_rate_hertz=44100,
language_code='en-US',
enable_word_time_offsets=True)
operation = client.long_running_recognize(config, audio)
print('Waiting for operation to complete...')
result= operation.result(timeout=6000)
# Each result is for a consecutive portion of the audio. Iterate through
# them to get the transcripts for the entire audio file.
for result in result.results:
alternative = result.alternatives[0]
print('Transcript: {}'.format(alternative.transcript))
print('Confidence: {}'.format(alternative.confidence))
for word_info in alternative.words:
word = word_info.word
start_time = word_info.start_time
end_time = word_info.end_time
print('Word: {}, start_time: {}, end_time: {}'.format(word,start_time.seconds + start_time.nanos * 1e-9,end_time.seconds + end_time.nanos * 1e-9))
if __name__ == '__main__':
gcs_uri="gs://speechmldemo/DirtyAudioExample.flac"
transcribe_gcs(gcs_uri)r code here
Here you could find a sample which the stereo file is split into independent channels in memory, then auditok tokenizes each channel based on values lower than a given threshold indicating silence. Each segment is sent to Google (preferably in parallel when their limits are better), and the responses are collected and sorted according to the time in which they occurred.