akasaki1211 / whisper2ae

Whisper to AE

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.

Setup

1. Create a venv (optional)
2. Install whisper

cd whisper2ae
python -m venv whisper
.\whisper\Scripts\activate

pip install git+https://github.com/openai/whisper.git 

1. Transcribe for AE

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

2. Load to AE

Run txt2aeLayer.jsx in AfterEffects and select the result txt. If a Text Layer is selected, it will be created by duplication.