every_holiday/VRCT
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge branch 'AudioTranscriber' into develop

Browse Source
This commit is contained in:
misygauziya
2023-08-11 01:53:28 +09:00
parent f58b08c5a4 115cf8fab9
commit 61837ae6de
2 changed files with 5 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
VRCT.py
View File

@@ -467,12 +467,11 @@ class App(CTk):
self.mic_transcriber = AudioTranscriber( self.mic_transcriber = AudioTranscriber(
speaker=False, speaker=False,
source=self.mic_audio_recorder.source, source=self.mic_audio_recorder.source,
language=transcription_lang[self.INPUT_MIC_VOICE_LANGUAGE],
phrase_timeout=self.INPUT_MIC_PHRASE_TIMEOUT, phrase_timeout=self.INPUT_MIC_PHRASE_TIMEOUT,
max_phrases=self.INPUT_MIC_MAX_PHRASES, max_phrases=self.INPUT_MIC_MAX_PHRASES,
) )
def mic_transcript_to_chatbox(): def mic_transcript_to_chatbox():
self.mic_transcriber.transcribe_audio_queue(self.mic_audio_queue) self.mic_transcriber.transcribe_audio_queue(self.mic_audio_queue, transcription_lang[self.INPUT_MIC_VOICE_LANGUAGE])
message = self.mic_transcriber.get_transcript() message = self.mic_transcriber.get_transcript()
if len(message) > 0: if len(message) > 0:
# word filter # word filter
@@ -566,13 +565,12 @@ class App(CTk):
self.spk_transcriber = AudioTranscriber( self.spk_transcriber = AudioTranscriber(
speaker=True, speaker=True,
source=self.spk_audio_recorder.source, source=self.spk_audio_recorder.source,
language=transcription_lang[self.INPUT_SPEAKER_VOICE_LANGUAGE],
phrase_timeout=self.INPUT_SPEAKER_PHRASE_TIMEOUT, phrase_timeout=self.INPUT_SPEAKER_PHRASE_TIMEOUT,
max_phrases=self.INPUT_SPEAKER_MAX_PHRASES, max_phrases=self.INPUT_SPEAKER_MAX_PHRASES,
) )
def spk_transcript_to_textbox(): def spk_transcript_to_textbox():
self.spk_transcriber.transcribe_audio_queue(self.spk_audio_queue) self.spk_transcriber.transcribe_audio_queue(self.spk_audio_queue, transcription_lang[self.INPUT_SPEAKER_VOICE_LANGUAGE])
message = self.spk_transcriber.get_transcript() message = self.spk_transcriber.get_transcript()
if len(message) > 0: if len(message) > 0:
# translate # translate

7
audio_transcriber.py
View File

@@ -9,9 +9,8 @@ PHRASE_TIMEOUT = 3
MAX_PHRASES = 10 MAX_PHRASES = 10
class AudioTranscriber: class AudioTranscriber:
def __init__(self, speaker, source, language, phrase_timeout, max_phrases): def __init__(self, speaker, source, phrase_timeout, max_phrases):
self.speaker = speaker self.speaker = speaker
self.language = language
self.phrase_timeout = phrase_timeout self.phrase_timeout = phrase_timeout
self.max_phrases = max_phrases self.max_phrases = max_phrases
self.transcript_data = [] self.transcript_data = []
@@ -27,7 +26,7 @@ class AudioTranscriber:
"process_data_func": self.process_speaker_data if speaker else self.process_speaker_data "process_data_func": self.process_speaker_data if speaker else self.process_speaker_data
} }
def transcribe_audio_queue(self, audio_queue): def transcribe_audio_queue(self, audio_queue, language):
# while True: # while True:
audio, time_spoken = audio_queue.get() audio, time_spoken = audio_queue.get()
self.update_last_sample_and_phrase_status(audio, time_spoken) self.update_last_sample_and_phrase_status(audio, time_spoken)
@@ -37,7 +36,7 @@ class AudioTranscriber:
# fd, path = tempfile.mkstemp(suffix=".wav") # fd, path = tempfile.mkstemp(suffix=".wav")
# os.close(fd) # os.close(fd)
audio_data = self.audio_sources["process_data_func"]() audio_data = self.audio_sources["process_data_func"]()
text = self.audio_recognizer.recognize_google(audio_data, language=self.language) text = self.audio_recognizer.recognize_google(audio_data, language=language)
except Exception as e: except Exception as e:
pass pass
finally: finally: