pip install deepl tqdm TTS faster-whisper imageio
- put video in vt/ori
- run v2srt.py
- modify auth_key = 'deepL_auth_key', put chinese srt file in vt/srt
- run srt_trans.py
- modify speakers example in vt/ori/speakers
- 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) - run srt2speech.py
- pip install spleeter
- run split_bg.py