fesvhtr/vt

Init

pip install deepl tqdm TTS faster-whisper imageio

video to srt

1. put video in vt/ori
2. run v2srt.py

srt translation

1. modify auth_key = 'deepL_auth_key', put chinese srt file in vt/srt
2. run srt_trans.py

srt to speech

1. modify speakers example in vt/ori/speakers
2. download weight from https://huggingface.co/coqui/XTTS-v2, modify init_xtts()
   set speaker and translated srt file: translate_all_srt(tts_model, speaker, srt_file)
3. run srt2speech.py

split background audio

1. pip install spleeter
2. run split_bg.py