dli7319/slides-voiceover-generator

Convert slides and script into a video with TTS voiceover

slides-voiceover-generator

Convert slides and script into a video with TTS voiceover.
This is a set of bash scripts to help convert a set of slides into a video with a voiceover.
The voiceover is generated using coqui-ai/TTS.
The scripts are designed to be run on a Linux machine with docker, an nvidia gpu, and ffmpeg installed.

Usage

1. Clone this repository
2. Add a file called script.txt and add slides.pdf.
   • script.txt should contain the script for the voiceover, one line per slide.
   • slides.pdf should contain the slides.
3. Run ./extract_slides.sh to extract the slides into individual images using pdftoppm.
4. Run ./generate_voiceover.sh to generate the voiceover.
5. You can now use any video editting software like DaVinci Resolve to combine your rasterized slides and voiceovers.

Utility scripts:

1. Run ./initialize_durations to initialize the durations of each slide.
   • You can tune durations.txt to your liking, e.g. for overlay videos.
2. Run ./stitch_video.sh to combine the slides and voiceover into a video.
   

Notes

• For stitch_video.sh, you can optionally an overlays.txt file to add video overlays to the video.
  This should be a txt file where each line has the format:
  filename;width:height;position_x:position_y;start_overlay_time;end_overlay_time;start_time
  For example, to put overlay.mp4 in the top-left corner from time 0s to 10s:
  overlay.mp4;200:-2:0:0;0:10:0