Remove and Combine Backgrounds with AI

AI Background Removal for A1111 and ComfyUI - With this powerful method you can remove and change backgrounds. This is very powerful for logos and gaming assests. But also to put characters into different Backgrounds.
#### Links from my Video ####
Get my Workflow here: / 101592266
github.com/AUT...
github.com/chf...
github.com/WAS...
#### Join and Support me ####
Buy me a Coffee: www.buymeacoff...
Join my Facebook Group: / theairevolution
Joint my Discord Group: / discord
AI Newsletter: oliviotutorial...
Support me on Patreon: / sarikas

[@BlueScorpioZA]

You can add Rembg directly to the TXT2IMG interface if you configure it in settings in Auto1111 or Forge. Settings - Post processing. Select "Rembg" in the Enable Postprocessing operations in txt2img and img2img tabs, save, restart and you're ready to go.

[@anonymousgamer6665]

Is rembg integrated with the A1111 forge interface? If so how to make it work?

[@joehogans4494]

Awesome thanks!

[@fabricemag4151]

Thank you for this video. An interesting solution is sd-forge-layerdiffuse. You can create the character and background separately and then combine them. ( among other things )

[@jcvijr]

I’ve used this extension for a while, you can batch video frames to remove bg 👍

[@brohemian]

any easier interface /solution for this?

[@jcvijr]

@@brohemian there are lots of easier interfaces/solutions who does it in the web, but they are all paid or limited for free use :)

[@brohemian]

@@jcvijrthx,

[@nolanzor]

The banging intros are awesome, keep doing them

[@OlivioSarikas]

Thank you :)

[@erdbeerbus]

this is indeed the greatest improvement I have seen using ai ... in relation to my work :) great Mr Sarikas!

[@MikevomMars]

Although I am a fan of ComfyUI, I don't like that over-complicated solutions to do things that can be done in a modern image editor with a single click only. So this time, I am about to prefer the A1111 way.

[@goyetus]

Hello Olivio! thanks for all your videos !!!

[@davidrosbo4519]

Check out layer difuse for removing backgrounds. Extension for txt 2 img

[@OlivioSarikas]

thank you, i will check that out

[@hleet]

I was looking for a way to do that , and voila 🎉. My first approach was to do a mask of the person and then invert it to inpaint the background, but it wasn‘t smooth enough. I think I‘ll do that in affinity photo instead 😅

[@electronicmusicartcollective]

Super introduction. Love you new individual intros. keep on doing this to make every episode more personal and where your colourfull shirts? Its your trademark. best wishes

[@OlivioSarikas]

Thank you :D

[@metaltech3944]

Great Stuff and I love the shades.

[@davidkerr7]

you can do this with A111 + ComfyUI extension u can run a workflow as a pre or post process on an image or latent

[@jeremyleonbarlow]

You can add rembg to post processing in automatic1111 and have it removed on creation.

[@I-Dophler]

Great explanation of how to remove backgrounds and mix tools like Automatic 1111 and Com YOUi! You showed how to use different add-ons and pick the best background color really well. And Mr. Butters, starting your videos with those rock-star glasses? Genius! They really make your videos stand out. Thanks for sharing; I can't wait to try these tips in my work.

[@Willtry-l6g]

Thank you for your videos😊 Can I know if, when we have 5 pictures for each of two different subject(assume a1 person and a2 person), using stable diffusion can we generate these tree situation in one model? 1.a1 person in different background 2. a2 person in different backgrounds 3. a1 and a2 in the same image with different background . Do I need to create 3 model for that task? How to do that type of task ?

[@sazarod]

Cool extension and all but, why not just generate the image with a transparent background to begin with? There is an extension for that

[@GaryStafford]

Thanks!

[@OlivioSarikas]

Thank you for your support :)

[@melchiorao9759]

Olivio have you tried the new BiRefNet comfy custom node? By far the best bg remove to date. Especially for video, where every frame is important.

[@Deeshu-qd5em]

I'm using image from outside using load image, while using Image Alpha composite I'm getting error - image size must be same. I used image resize to make sure I pass same size of both image but getting same error again. Any solutions please?

[@luiscaro3]

I'd like to know how this compares to the new stuff that Photoshop is putting out. Right now you can get PS for thirty something bucks a month and it's pretty impressive and easy to use, thoughts?

[@BabylonBaller]

haha INtro is hilarious, We miss Hacker Man that wig!

[@daviddiehn5176]

Thank you!

[@alwardslab8732]

Used rembg for some time but had to photoshop in post the results, dark borders, missing masked areas, hair/fur uncanny cuts... layer diffuse is way way cleaner... but layer breaks with ipadapters (i used them for cosistency)

[@henroc481]

What should I look into for an ambient light layer overlay in comfy?

[@DivinityIsPurity]

A1111? I'm watching this one

[@OlivioSarikas]

yes, A1111 is covered too :)

[@erdbeerbus]

everything is running till my issue: No such file or directory: '/Users/admin/tutorial/ComfyUI/custom_nodes/ComfyUI_LayerStyle/RMBG-1.4/model.pth' ... but I am sure to manage this ... maybe :)

[@erdbeerbus]

mmmm, I assume the issue is the BRIA loader ... anyone here with the same problem?

[@Amelia_PC]

Hmm, maybe it works for animation, adding backgrounds in image batches with characters only. It'd be better than rendering it directly or spending money on a DaVinci Resolve Pro license.

[@OlivioSarikas]

yes, should work too. Also check the magic mask in davinci resolve. Because SD will change the details on your character and is also not made for longer video sequences

[@Amelia_PC]

@@OlivioSarikas Thanks! But magic mask is only available on Pro version. "Because SD will change the details on your character and is also not made for longer video sequences" Well reminded! I forgot about these "little details" (that changes everything in animation). So it's best to stay on Blender and 3ds max, rendering the background in a plane with material when the background is not 3D.

[@zinheader6882]

nice glasses, is that AI too?

[@morozig]

Hey, @OlivioSarikas, thanks for all your videos! Can you please make review of PonyXL model? I think it's the most misunderstood model of all.

[@Gifttv204]

Please help me 😢 how to fix, samplemodel error on comfy 😢help me please,i can't able to use comfy

[@xnoexe]

is it still working?

[@Sebastian78120]

So we can remove the bg in A1111 but we can't change it like in Comfy ?

[@OlivioSarikas]

not really, since there is no way in A1111 to bring together different Ai images in a automated way

[@joe.todddq]

I miss your blonde wig

[@forfreeiran8749]

We need more videos using forge maaaan please

[@Elwaves2925]

The section on A1111 would also cover Forge and the same would apply to other videos too.

[@LouisGedo]

👋

[@sb6934]

Thanks!