Capgen is an automatic captions generator for videos. It employs Whisper neural network, offered by OpenAI to generate accurate timestamped captions for your videos in srt file format. Whisper is introduced here, https://openai.com/blog/whisper/. The code for Capgen is a from-scratch re-implementation of the code released by OpenAI. Support for other output formats will be added in the futuređź‘€.
- ffmpeg cmd program for extracting audio from video available at https://ffmpeg.org/download.html
- Download the source code.
cdto capgen\capgen directory.- Download and install ffmpeg cmd.
- Run
pip install -r requirements.txtto install dependencies.
- Run
python capgen.py <path-to-video>. An srt file is generated in the same directory. - Run
python capgen.py -hto see all the available options.