Added seed travel and img2img video to the wiki.

[ClashSAN]

Really like the video ones and will be looking for more related scripts!
seed travel I made:
https://user-images.githubusercontent.com/98228077/192487018-06e1b86a-8763-4e8c-906b-273754f161cd.mp4

The demo video of shrek walking:
https://user-images.githubusercontent.com/98228077/192490706-a796894e-8a5a-4ab5-bf15-846ba8d3024c.mp4

There are a couple of scripts I found mentioned worth testing:
Getting Textual Inversion Embeddings:
#757

4chan animation scripts - referenced here: #792 (comment)
Animation Script v0.2
https://pastebin.com/h3qXa8hZ

Animation Script v.02 with log modulation, sine modulation, and parameter reset frames.
https://pastebin.com/QZc3dnPx

Credits:
https://github.com/yownas/seed_travel
https://github.com/memes-forever/Stable-diffusion-webui-video

[Filarius]

Maybe you will be interested in code of video2video who do not unpack frames to hard drive
https://github.com/Filarius/stable-diffusion-webui/blob/master/scripts/vid2vid.py (later i will move it to separate repo).
But it do not do anything beside simple img2img per frame.
Looking for help to add latent interpolation here.

[ClashSAN]

Updated, ty
https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Custom-Scripts

[sbaier1]

FWIW out of curiosity and as a learning experience i tried interpolating the latent representations after diffusion sampling before the VAE of two prompts to see what slerp interpolations look like at that level

in itself it isn't very interesting yet, but perhaps adding intermediate diffusion runs with weighted prompts and then interpolating between those steps might be fun 🤔 or maybe combining this with img2img alt so the modifications to the image can be visualized in a smooth diff

(am also generally interested in interesting video scripts/ideas for sampling the model)

master...sbaier1:stable-diffusion-webui:feat/prompt-interpolation

interp_test-2x-RIFE-RIFE3.1-36fps.mp4