Transcribe from a video with audio using OpenAI Whisper and turn the result into a Text Layers in AfterEffects. Text Layers are split into segments.
- Create a venv (optional)
- Install whisper
cd whisper2ae
python -m venv whisper
.\whisper\Scripts\activate
pip install git+https://github.com/openai/whisper.git
The following command will download the models into the models
folder and begin transcription. The result txt
file will be created in the outputs
folder.
Minimum command :
python whisper_transcribe.py --path .\sources\input.mp4
with options :
python whisper_transcribe.py --path .\sources\input.mp4 --model large --language ja --savejson 1
Run txt2aeLayer.jsx in AfterEffects and select the result txt
. If a Text Layer is selected, it will be created by duplication.