This is just an idea I do not know where to share.
From a small B/W video/GIF it is possible to create ASCII frames.
Something like this:
Get video frames as PNG.
ffmpeg -i video.mp4 -vf fps=10,crop=800:300:400:100,scale=80:-1,negate,eq=contrast=5,format=gray frames/output_%04d.png
Convert to ascii frames:
for img in frames/*.png; do
./mediatoascii --image-path "$img" > ascii_frames/$(basename "$img" .png).txt
done
Show in terminal:
for frame in ascii_frames/*.txt; do
clear
cat "$frame"
sleep 0.1
done
Just an example. I'm pretty sure this can be improved a lot. For instance, there are some ASCII to SVG tools so maybe ascii could be optimized for SVG rendering.
So that, the repository could host video prompts and ascii rendered frames. Rendered frames could be easily improved via PR. AI could probably recreate video from the frames in a near future.
This is just an idea I do not know where to share.
From a small B/W video/GIF it is possible to create ASCII frames.
Something like this:
Get video frames as PNG.
Convert to ascii frames:
Show in terminal:
Just an example. I'm pretty sure this can be improved a lot. For instance, there are some ASCII to SVG tools so maybe ascii could be optimized for SVG rendering.
So that, the repository could host video prompts and ascii rendered frames. Rendered frames could be easily improved via PR. AI could probably recreate video from the frames in a near future.