π° About
I previously wrote an article that showed, how, for totally free of charge, and with 100% open source software, it is possible to analyze photos and produce textual analysis:
![adriens](https://media2.dev.to/cdn-cgi/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F446871%2F3e9ded5c-f368-4906-a277-35e56c9f97a7.png)
πΈ Disaster reporting w/Open Source LLaVA/AI for photo analysis
adriens γ» May 21
π Today, I'll show, what can be achieved on top of these textual analysis.
π What we'll do
This time, we will enrich the previous pipeline with the following workflow:
-
π¨
text2img
: inject text analysis into an image with https://ideogram.ai -
π¬
img2video
: Put resulting image into Gen-2 by Runway to get a movie clip with - ποΈ Prepare a small storytelling
- πΆ Select a soundtrack
- ποΈ Put all the stuff into a single short movie clip
Enough talk, here is the resulting movie clip.
ποΈ Demo
Top comments (1)