Hacker News

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

123 points by DSemba ago | 29 comments

james2doyle |next [-]

There are some demo spaces using this. This one seems the best (paint your own mask) but it failed on all the images I tried: https://huggingface.co/spaces/multimodalart/Moebius

delis-thumbs-7e |next |previous [-]

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

doctorpangloss |root |parent [-]

how many times have you edited a photo you took on your phone in the last 7 days?

stusmall |root |parent |next [-]

I think 3? I feel like that's often enough. Sometimes it's nice to do a quick dumb ass gag on a whim. If I am anything I am a man who loves a dumb ass gag.

TeMPOraL |root |parent |next |previous [-]

Half a dozen at least.

(I'm counting only times I used generative editing options in my Galaxy phone - if I were to take your question literally, it would be "at least once every other day", simply due to rotating and cropping.)

dogomatic |root |parent |previous [-]

Personally, about 9 times. Would be higher if it was even easier and cheaper

hari1123 |next |previous [-]

lot of the photo editors on mobiles have this, maybe even some apps?

gspr |next |previous [-]

Nitpick: in the showcase on that page, under Comparison of Natural Scenes, Moebius should definitely get a "structural confusion" tag for the back of the surfboard. If other models get deducted for truncating the surfboard, then surely the elongation that Moebius does should count too.

Also, what's going on behind the in-painted corner of the house? We'd need to see higher resolution pictures, but I'm not convinced that it too shouldn't get a flag. Likewise with the beach just behind the surfboard. Not terrible, but what gets flagged in the competitors is similar.

NooneAtAll3 |next |previous [-]

I don't understand. Is it available somewhere to try or is it just an ad?

owebmaster |root |parent [-]

Yeah it's great but how do I use it?

Edit: I think I found it https://huggingface.co/hustvl/Moebius

K0IN |root |parent [-]

with this size we could have a interaactive web demo.

GL26 |next |previous [-]

Could this run locally on a smartphone ?

teroshan |next |previous [-]

Unrelated but when I read inpainting and Moebius I was scared it was related and using the art of the great Jean Giraud [0] a.k.a. Moebius

https://characterdesignreferences.com/artist-of-the-week-3/m...

[0] https://en.wikipedia.org/wiki/Jean_Giraud

coldtea |root |parent [-]

Scared why?

teroshan |root |parent [-]

Scared for the same reason I found last year's 'Ghibli filter' craze upsetting, I would have personally hated to have seen this artist's legacy used for promoting AI image generation.

TeMPOraL |root |parent [-]

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.

solid_fuel |root |parent [-]

> In case that happened then the rest of the world would probably appreciate the art

What art?

We’re talking about generated pictures, aka slop, not art made by a real human.

And I don’t know if you’ve been paying attention but people seem to be pretty tired of the slop. I don’t think it would be appreciated nearly as much as you think.

TeMPOraL |root |parent [-]

This definition of "slop" doesn't cut reality just quite at the joints.

People are tired of marketing. AI generated slop people are annoyed with, is garbage produced for marketing reasons, and it's distinctly noticeable precisely because all the bottom-feeder marketing houses switched to using it. But it's not the AI itself that's the problem here. Slop was here before, but it was made with cheap protein-based image generators. Silicon-based generators are just cheaper.

rasz |next |previous [-]

It sure has a thing for chins, jaws and removing weight, looksmaxing build in.

N_Lens |next |previous [-]

The gallery of their samples is pretty impressive!

epolanski |next |previous [-]

What is the current SOTA for impainting?

I have a potential project for my e-commerce where I want to allow users to upload images of their house exteriors and impaint awnings.

TeMPOraL |root |parent |next [-]

Awnings, if I understand correctly (I just learned this word right now), are purely additive attachments to structure exteriors - so perhaps they wouldn't necessarily need a full inpainting model? Wouldn't it be enough to estimate an affine transform for a quad and blend the image of awning directly (and the same with shadow map to fake shade)? Is classical photogrammetry up to such task these days?

jdiff |root |parent [-]

I'm quite perplexed by this comment. If I'm understanding you correctly, sure, what you describe is possible through significantly more effort, orchestration, and source photos. Or we can grab one still image and throw an inpainting model at it.

vunderba |root |parent |next |previous [-]

Proprietary? Either gpt-image-2 or NB2.

I have an example of interior decorating inpainting where I replaced a large floor-to-ceiling window with a mirror, and the result was pretty impressive using NB Pro from nearly a year ago.

https://imgpb.com/ZXkiXV

Locally hostable? For my money I'd argue Flux.2 Klein but Qwen-Edit still puts in the work.

CharlesW |root |parent |next [-]

NB2 means "Nano Banana 2", a Google image generation model. https://blog.google/innovation-and-ai/technology/ai/nano-ban...

IAmGraydon |root |parent |previous [-]

As far as I know, gpt-image-2 doesn't even let you define a mask unless you've already run it through one iteration, and once you do define the mask, it just ignores it 90% of the time. It's utterly useless for inpainting. Also, this and other proprietary models are severely limited in their output resolution.

I do agree, however, that the Flux2 family is the SoTA at the moment. Running locally via something like Comfy gets incredible results.

vunderba |root |parent [-]

[dead]

BoredPositron |root |parent |previous [-]

flux klein with LoRa. GPT image and nano often produce high frequency artifacts when editing.

zb3 |previous [-]

1) What are RAM requirements?

2) If these are reasonable, a WebGPU demo would be great..