NEW! ControlNet 1.1 No Prompt Inpainting.

Ah, some lnet so I can f-off and no installation of that is being specified. Nice.

You really have NO idea what you're doing here. Your so called 'examples' are nonsensical.

What if I remove a logo from an outfit and want to upload an image to replace

Can you please help with the case when I want to put on my wall tiles, so that they will be right size and with perfect cells?

Thanks for the video😉

I see in your UI a calc, which extension is it?

Very nice!

Hello, thanks for the interesting video, I have a couple questions about it:
1 Did you compare the inpaint model controlNet and the Deliberate inpanit model? If you did, which one is preferable to use?
2 Do I understand correctly that the inpaint model controlNet is created for models that don't have their own inpaint version?

impaint_only,impaint_only+lama,inpaint_global_harmonious - what are differencies between them?

Did you know that the masking tools only work right with the Inpaint model? If you try to use it with any other model (to mask out part of your input image, for example), when you draw a mask, it dumps your image and only honors the mask you drew. In other words, if you draw a square, it will make an entirely new image, except where you drew the square. I wrote it up and they said it wasn't a bug... *shrug*. But what if I use the Inpaint model to mask in loopback mode? hmmm...

👋

I dont understand, how is this controlnet inpainting better than regular inpainting?

Did you ever stop singing wonder wall?

Thank you very helpful, is there any tutorial to upload mask image, instead of drawing it?

Can you use depth information out of an image and then do an inpaint? For instance a tattoo on the nose, on the ear, in the neck going up to your cheek? Thanks man!

maybe you have to use a prompt?

I had assumed inpainting did this already but now can see why I struggled with it sometimes.

I have "FileNotFoundError: [Errno 2] No such file or directory: 'C:\\New Folder\\tmp\\gradio\\" error, what's can be wrong?

Depth gave better results for me

Thanks for such a useful video. but I don't have Control Type buttons in my Control net section and control net is in 1.1.150 version !!!! i updated it for 5 times. but no changes. could you me help Mr Kampth ? why is that ? Thanks

wow i try to use, thank you for example👍

Fantastic!!!

Here's a scenario I think might be quite helpful, trying to get rid of an item completely, e.g. say you have a portrait of a woman wearing and she's wearing a wristband and you want to get rid of the wristband. With normal in painiting it can be quite an effort to convince it to get rid of an item and replace it with something that looks realistic. This might work better because it understands the context but doesn't derive anything from the place where you're getting rid of it.

It seems that ControlNet replaces the Stable diffusion

Do prompts not work? It's kind of useless to me if they don't.

uhm... so I've been wondering... about that mid journey thing... why the heck would you want to pay to use a discord bot? is it for people who dont have a pc with VRAM?... and can I just rent out my video card?

seems you didn't do your homework before making this video, bad results

Still not as breakthrough as Firefly though, maybe it will be as good as Firefly if we give the ControlNet's developers some time

Did you've already trying aedetailer extension? It seem smart emough to fix Eyes and hand problems in low resolution or mid shot type render without bother with hires fix

I believe It also use inpainting method but work automatically

As far as I understand, the point of the inpainting ControlNet is that it works with non-inpainting models. You can take any model and inpaint more efficiently without having to retrain or do a hacky inpaint model merge.

Can it make portrait photo became landscape photo like Generative Fill in Photoshop beta?

Came for your guides and updates about stable diffusion, stayed for your Dad jokes/puns 😂

I couldn't really see anything different, on the contrary, why use luck for that if it can be directed towards a better solution? or were these examples not cool..

how do you have the controlnet units? is it a separate extension?

nice video i love your experiments ! but controlnet inpainting is not very good at the moment I think

Is there actually a realistic vision inpainting model or is it just a merge of realistic vision and 1.5 inpainting model?

The results were... okish? I didn't see anything that justifies that control net "Inpainting Just got Better!" as the title says.... mmmm

Why would you use deliberate_v2 for inpainting, and not Deliberate-inpainting?

Nowhere near as good or as fast as Photoshop. Hopefully it will improve in the future, but for now the best AI inpaint/outpaint is Photoshop Firefly

I wonder if this can be used for out painting. Extend the image and leave the new area black and see what happens...

Why do you increase batch count rather than batch size?

Commercials just too much! Completely ruin video.

very good results with the protogen models

Great update!

I'm sure there's a place for this, but I honestly don't see this producing results better than what you'd get with normal inpainting when used with a good prompt. Moreover, the addition of the Photopea extension now means we have precise masking control as well as photo bashing when doing inpainting, which is such an amazing combination.