I've been monitoring some of the image generation progress for game sprites for a while and was really impressed by Gemini Flash. This demo lets you upload a picture and turns it into an animated pixel character (a lil digi gif that you can download). I chain together the pixel art frames and insert the player into a tricky platforming level.
video demo:
https://www.youtube.com/watch?v=EK7z-HltssE
api usage:
Unfortunately Gemini Flash 2.0 image Gen has really tight API limits that don't scale up for paid tiers -- if anyone on the team sees this and wants to help me out, my email is in my profile or you can reach me at hello@lildigi.me
level stats:
For the speedrunners out there, the fastest glitchless times I've seen around 17-18s.
stack:
cloudflare pages + aws lambda + gemini
happy to answer any questions!
adamddev1 1 day ago
This is one of the most delightful things I've seen on the internet in a long while. I immediately started sharing it with my friends and we shared the cute little characters we made. Everyone loved it.
sunnyba 1 day ago
Thank you, I appreciate you sharing that!
IanCal 1 day ago
That's a fantastic idea! I love the little me on that - taken from just a webcam pic with a messy background and it's pulled out things like my tshirt perfectly.
sunnyba 1 day ago
Thanks! A few friends have starting using the generated digis as profile pictures because of how recognizable they are. I was definitely surprised at how good the models had gotten.
armitage__ 1 day ago
Very neat.
Would you share some of the prompts you used to generate the pixel art sprite? Super accurate to the photo I provided.
sunnyba 1 day ago
Thanks! Like another user commented below, I think the appeal of the pixel art sprites is that at a lower level of detail it can feel really accurate. 10 different real life red jackets can all end up as the same pixel art representation but each person would recognize it as their jacket! It feels like a form of compression.
On prompting, you can get most of the way there in AI studio on Gemini 2.0 Flash (Image Generation) Experimental by uploading a picture and asking for "a high quality detailed pixel art sprite of this character." Most of the backend annoyance here was iterating to improve prompt adherence (characters not facing the same way, outfits changing between frames, etc).
apgwoz 1 day ago
I am extremely impressed. I literally put my hand fully in front of my face, and it got me spot on. My glasses were partially shown, as was my beard and hair.
That being said, the resolution is such that saying messy hair, full beard, and black glasses, would pretty much get it.
tough 1 day ago
He's probably feeding the webcam pic as img2img no? That'd get you the most details from the original picked up by the AI
apgwoz 1 day ago
I uploaded a photo, though. I didn’t give permission for the camera.
I think it’s simply a matter of it assumed a person, found glasses, and a bit of hair above and below and filled in logically.
sunnyba 1 day ago
thanks for the traffic everyone -- definitely hitting rate limits on gemini already so if it doesn't work please give it a try later in the day or try the example.
Apologies for the inconvenience! If anyone on the gemini team can help out, please lmk (email in bio).
