Update app.py
Browse files
app.py
CHANGED
@@ -35,6 +35,10 @@ import torch
|
|
35 |
from TTS.tts.configs.xtts_config import XttsConfig
|
36 |
from pydub import AudioSegment
|
37 |
from pyannote.audio import Pipeline
|
|
|
|
|
|
|
|
|
38 |
|
39 |
# Accept license terms for Coqui XTTS
|
40 |
os.environ["COQUI_TOS_AGREED"] = "1"
|
@@ -157,7 +161,7 @@ def segment_background_audio(audio_path, output_path="background_segments.wav"):
|
|
157 |
result_audio.export(output_path, format="wav")
|
158 |
print(f"Saved non-speech (background) audio to: {output_path}")
|
159 |
|
160 |
-
return
|
161 |
|
162 |
def transcribe_video_with_speakers(video_path):
|
163 |
# Extract audio from video
|
@@ -166,7 +170,7 @@ def transcribe_video_with_speakers(video_path):
|
|
166 |
video.audio.write_audiofile(audio_path)
|
167 |
logger.info(f"Audio extracted from video: {audio_path}")
|
168 |
|
169 |
-
segment_background_audio(audio_path)
|
170 |
print(f"Saved non-speech (background) audio to local")
|
171 |
|
172 |
# Set up device
|
@@ -343,7 +347,6 @@ def update_translations(file, edited_table, mode):
|
|
343 |
except Exception as e:
|
344 |
raise ValueError(f"Error updating translations: {e}")
|
345 |
|
346 |
-
|
347 |
def create_subtitle_clip_pil(text, start_time, end_time, video_width, video_height, font_path):
|
348 |
try:
|
349 |
subtitle_width = int(video_width * 0.8)
|
@@ -484,14 +487,6 @@ def add_transcript_voiceover(video_path, translated_json, output_path, add_voice
|
|
484 |
|
485 |
return error_messages
|
486 |
|
487 |
-
import os
|
488 |
-
import traceback
|
489 |
-
from TTS.api import TTS
|
490 |
-
import wave
|
491 |
-
import logging
|
492 |
-
|
493 |
-
logger = logging.getLogger(__name__)
|
494 |
-
|
495 |
# Initialize TTS model only once (outside the function)
|
496 |
tts = TTS(model_name="tts_models/multilingual/multi-dataset/xtts_v2")
|
497 |
|
|
|
35 |
from TTS.tts.configs.xtts_config import XttsConfig
|
36 |
from pydub import AudioSegment
|
37 |
from pyannote.audio import Pipeline
|
38 |
+
import traceback
|
39 |
+
import wave
|
40 |
+
|
41 |
+
logger = logging.getLogger(__name__)
|
42 |
|
43 |
# Accept license terms for Coqui XTTS
|
44 |
os.environ["COQUI_TOS_AGREED"] = "1"
|
|
|
161 |
result_audio.export(output_path, format="wav")
|
162 |
print(f"Saved non-speech (background) audio to: {output_path}")
|
163 |
|
164 |
+
return True
|
165 |
|
166 |
def transcribe_video_with_speakers(video_path):
|
167 |
# Extract audio from video
|
|
|
170 |
video.audio.write_audiofile(audio_path)
|
171 |
logger.info(f"Audio extracted from video: {audio_path}")
|
172 |
|
173 |
+
segment_result = segment_background_audio(audio_path)
|
174 |
print(f"Saved non-speech (background) audio to local")
|
175 |
|
176 |
# Set up device
|
|
|
347 |
except Exception as e:
|
348 |
raise ValueError(f"Error updating translations: {e}")
|
349 |
|
|
|
350 |
def create_subtitle_clip_pil(text, start_time, end_time, video_width, video_height, font_path):
|
351 |
try:
|
352 |
subtitle_width = int(video_width * 0.8)
|
|
|
487 |
|
488 |
return error_messages
|
489 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
490 |
# Initialize TTS model only once (outside the function)
|
491 |
tts = TTS(model_name="tts_models/multilingual/multi-dataset/xtts_v2")
|
492 |
|