Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Moebius: 0.2B image inpainting model with 10B-level performance

hustvl.github.io

Post reply on HN
Print view View original

71–80 of 86 posts

Re: Moebius: 0.2B image inpainting model with 10B-level performance

#71

Post by epolanski »

TeMPOraL wrote: post #54
epolanski wrote: post #36

Earlier quoted context omitted.

I have no idea but I think you might be onto something. So you're saying that, if I can calculate from the picture the position (height, inclination and such), and I can render the model (should be doable) for that height and angle, my best course of action could be to combine original + render and only at the end use a visual model? That could be interesting.

More-less, yes. I was actually thinking about taking a high-resolution rendering of an awning directly facing the camera, and transforming that quad onto the user image - which requires computing the transformation matrix that would right the user image so the building is level and directly facing the camera, and then applying the inverse of that to the quad with your rendering - but I now realize I assumed user phot…

Yes, that's what we need to add (the red awning) to user provided pictures.

https://imgur.com/a/Y0Q4mfu

By the way I have a tried a handful of NB2 queries providing a reference image of the awning and of a user uploaded garden-facing building and I was very impressed by the results, I think that combining a 3d render at the right angle of the awning + NB 2 should do great.

Thanks a lot for your help, your feedback has been crucial! Dziekuje!

Re: Moebius: 0.2B image inpainting model with 10B-level performance

#72

Post by somenameforme »

zerobees wrote: post #40
delis-thumbs-7e wrote: post #18

This is the useful AI stuf. There’s so many usecases this makes possible.

Right, and that's what I find frustrating. There are so many use cases where a local, purpose-built model that's dependably good at one thing would really make a difference. But no one is going to throw a billion dollars to give us amazing dust removal, flawless scene segmentation, etc. Instead, you're supposed to upload it to the cloud and ask a big, multimodal frontier model to maybe please do the thing you want an…

You can do all of this locally on a cheap video card. Search for fooocus or automatic1111 for a couple of setups that are fairly low friction to get going. Amuse AI is another one. It's not quite state of the art and also censored, but it's by far the least friction (especially if you have an AMD card) - it's pretty much plug and play. ComfyUI is the advanced do-everything workhorse. However, it's anything but comfy if you don't already have a lot of knowledge about this domain. I'd generally recommend fooocus for a balance between usability and power/flexibility.

The million image gen services online are mostly just making bank off ignorance. People don't realize that their own cheap video cards are more than enough to do everything they're paying a service an orders of magnitude markup for.

Re: Moebius: 0.2B image inpainting model with 10B-level performance

#73

Post by Someone »

zerobees wrote: post #40
delis-thumbs-7e wrote: post #18

This is the useful AI stuf. There’s so many usecases this makes possible.

Right, and that's what I find frustrating. There are so many use cases where a local, purpose-built model that's dependably good at one thing would really make a difference. But no one is going to throw a billion dollars to give us amazing dust removal, flawless scene segmentation, etc. Instead, you're supposed to upload it to the cloud and ask a big, multimodal frontier model to maybe please do the thing you want an…

> There are so many use cases where a local, purpose-built model that's dependably good at one thing would really make a difference. But no one is going to throw a billion dollars to give us amazing dust removal, flawless scene segmentation, etc.

iPhones have models for text extraction and in-painting in the Photos App.

Both don’t have knobs to tune them, but, I think, they are decent for their intended audience (definitely not flawless, but I don’t think that exists anywhere, even if dropping the ‘local’ requirement)

For scene segmentation, iOS has models for detecting persons (https://developer.apple.com/documentation/Vision/segmenting-...).

It also has models for detecting faces, face features, body and hand poses, or for picking the ‘best’ selfie from a set.

(And dust removal is fairly niche compared to these, I think. Or do I overlook some common use case for it that many people want?)

Re: Moebius: 0.2B image inpainting model with 10B-level performance

#74

Post by bilekas »

zmgsabst wrote: post #64
teroshan wrote: post #53

Earlier quoted context omitted.

Perhaps. I like the idea that a piece of art, in addition of ultimately ending up as pixels on my screen, is also a window into a world that has been dreamt up by real human imagination, driven by their hopes and fears. Semiconductors based generation may give me the first part, but not the second. I'm speaking for myself here, I agree with your point though.

You just refuse to see certain people’s hopes and fears because they didn’t express them in a way you personally find pleasing. The LLMs didn’t prompt themselves.

> The LLMs didn’t prompt themselves.

I refuse to accept that real humans believe prompting is art.

Re: Moebius: 0.2B image inpainting model with 10B-level performance

#75

Post by NooneAtAll3 »

torgoguys wrote: post #67
chatmasta wrote: post #66

What is inpainting? Everyone in the comments seems to be familiar with the term, and I don’t see it described in the linked page.

Click on the visualizations to see it in action. The purple areas are areas a user highlighted to tell the system to inpaint, and when you click on the image you see the results of the inpainting. Basically the model redraws sections of an image (the purple areas) using the context of what's in the non-purple areas to decide what might look best in the purple areas. Often used for removing objects but as you can see…

> and when you click on the image

ah, bad UX

Re: Moebius: 0.2B image inpainting model with 10B-level performance

#76

Post by da_grift_shift »

g58892881 wrote: post #69
simonw wrote: post #61

I got this working with ONNX (thanks, Claude Opus 4.8) and now I have an interactive demo of the model running entirely in the browser here (~1.3GB download): https://simonw.github.io/moebius-web/ - code here: https://github.com/simonw/moebius-web (Claude Code transcript: https://gisthost.github.io/?58039ba5c1ca3ed177e8659168996ee4 ) Wrote this up in more detail on my blog: https://simonwillison.net/2026/Jun/22/porti…

well done! unet weights are in fp32. did you by any chance try something lower, fp16?

The model considered it.

There are 25 or so mentions of fp16 and fp32 weights across the 7500+ words of Markdown text it generated. So the next question might be: Did it make the right calls?

https://github.com/simonw/moebius-web/blob/main/notes.md

https://github.com/simonw/moebius-web/blob/main/plan.md

https://github.com/simonw/moebius-web/blob/main/research.md

https://github.com/simonw/moebius-web/blob/main/understandin...

Re: Moebius: 0.2B image inpainting model with 10B-level performance

#77

Post by NooneAtAll3 »

teroshan wrote: post #53
TeMPOraL wrote: post #16

Earlier quoted context omitted.

In case that happened then the rest of the world would probably appreciate the art, and a subset of it, the artist (and even a small subset of ~whole Internet-connected population is a lot of people ). Some silver lining, perhaps.

Perhaps. I like the idea that a piece of art, in addition of ultimately ending up as pixels on my screen, is also a window into a world that has been dreamt up by real human imagination, driven by their hopes and fears. Semiconductors based generation may give me the first part, but not the second. I'm speaking for myself here, I agree with your point though.

> I like the idea that a piece of art, in addition of ultimately ending up as pixels on my screen, is also a window into a world that has been dreamt up by real human imagination, driven by their hopes and fears.

I guess this actually defines the fringe between ai-art enjoyers and haters - some people prefer what art does to their imagination, while others look at what art does to others'

Re: Moebius: 0.2B image inpainting model with 10B-level performance

#78

Post by Yokohiii »

xrd wrote: post #41

I did an inpainting project for a client a few years ago. They were trying to inpaint banner ads for concert promoters, and find a way to make it easy to produce a bunch of different sized ads for a variety of placements. I was tasked with inpainting Xmas themed ad for a few major singers. The weirdest thing was when the inpainting tool added strange people to an image. This singer was all decked out in tinsel and re…

> At the time this was Stable Diffusion on the backend

The community made models (merges, fine tunes, etc) of that era are all completely overtrained and optimized for portraits and frontal shots. They would try to make a person out of anything. Inpainting faces is already a chore, even with a lot of tooling around that, but inpainting anything else is almost impossible. These models are also especially bad to fit objects naturally into scenes. You can make a crappy necklace or belt work, but introducing a new object into a scene just fails with infinite variety.

They are also much better using 512x512 as resolution, any larger deviation introduces more problems.

Considering you wanted to inpaint banner ads, they would probably get distorted heavily. Those models can't deal with fonts and are bad at a pixel perfect transfers. The only viable way to do this, at that time, would be to manually insert the banner ads and fix the seams with AI. Requires some artistic skill of course.

Your attempt was bold, but with the expectation of just supplying two images and let the models do it, it was impossible.

Re: Moebius: 0.2B image inpainting model with 10B-level performance

#79

Post by Yokohiii »

zerobees wrote: post #40
delis-thumbs-7e wrote: post #18

This is the useful AI stuf. There’s so many usecases this makes possible.

Right, and that's what I find frustrating. There are so many use cases where a local, purpose-built model that's dependably good at one thing would really make a difference. But no one is going to throw a billion dollars to give us amazing dust removal, flawless scene segmentation, etc. Instead, you're supposed to upload it to the cloud and ask a big, multimodal frontier model to maybe please do the thing you want an…

I have the feeling that the cloud based providers are just using the freely available segmentation models. It's just speculation, but it doesn't seem to be top priority for them, so they'd just bolt on anything that works.

A problem is also that the cloud solutions need a complex UI to surface segmentation to the user. But the point you have there is that those models are probably not prime time ready yet, surfacing them would actually reveal they are not as powerful as the user expects. Destroying the illusion that AI can just do anything at will.

Post reply on HN