I am a new subscriber. This is really a nice tutorial. Can you please make a tutorial using control net for poses. I want to make a person with one outfit (along with background) to match a pose of another person. Kindly make a step by step method. Also how to do in batches.
Hello thank you for the nice video! As I followed along, I got curious. Is there any way to put a picture of the clothes I want to wear instead of the clothes generated by AI?
Good one! Another way to replace clothes: Extract the pose from the same image (ControlNet Pose Full), use multi ControlNet with Pose: full without Preprocessor selected and only with pose image and model loaded + Depth (same here: without preprocessor, with model and extracted depth image loaded)+ Reference only (pick a photo with the cloth you want to use, it can be even a front view image of the cloth) Ref Only with "ControlNet is more Important" and 1.2 strength. Now you can use any cloth you want. It doesn't give 100% consistency (no AI model gives us this), but it kinda works.
@@RahulR-nm3kd Sorry! Cain't ^^ I forgot to mention that you need to make the changing on the Inpainting tab. (sample of what was done: i.imgur.com/pPotMh8.jpg) If you have any specific questions about something above, feel free to ask :)
hello brother. We keep it .5 to get a balance. If we keep it 1 then the output will be almost same as the original, if we keep it 0 then the output will be crazy different than the original
Nice video bro. Can we use something else other than depth? like softedge of canny? ((anyway, at first glance I really thought your CPU is burning in the background. LOL))
Thanks for the video. One question, I have a model that I generated in Midjourney and I want to dress her with an SPECIFIC dress (from a brand) it has to be exactly the dress but in the AI model I created. Is that possible using Stable Difussion (I'd add a photo of the dress in HD, white background) Thank you, I really need to solve this for a client! And at the moment I am photoshopping it. I think there must be softwares to do this
Hi just subscribed to your channel. and just finish talking to Canva Support of their fashion Ai clothes change. Anyway I was just at Dall-e and your channel just came up. Are you a photographer or a designer? or a Software engineer or just a content creatoretc... There is not much Niche on how hard to get a model, money, time etc.. and make the clothes perfect, my friend always complain about these... I mean there is some channel and website, but yours is particular. Cant wait to see this channel go viral.
Can we get same outfit in every photo? For example Japan kimono style, white color+pink Obi etc. I want exactly same outfit and ornaments in all photo.
Hey few questions. 1 Which version r u using 1.5 or 2.1? 2 I am not able to choose style. What could be the reason? Cant generate image also. Nothing comes up after clicking generate. I am using 2.1. 3 Where to find/install control net. I dont have it in mine.
bro please share the link! ki vabe bujhbo kon software ba kon website? ei tuku guide na korle next video gulo follow kori ki kore boss! guide us properly bro
it is not website feature, to use stable diffussion you need a pc with good graphic card and then you need to install stable diff step by step (you can find on youtube)
Constructive Feedback: 1. Background noise in the video can be removed using NVIDIA RTX voice application (also works with GTX cards) 2. Edit could be short as people will see content which is short without long pauses (it is engaging) 3. In the process we could avoid the face process is important which also allows us to just record a voice over Nice video !!
Cool tech! BTW kahma can create portraits of passed family, amazing.
Really one of the good tutorials on YT.
Thanks a lot
This was really helpful, thanks! 👍
No worries. Glad to assist
prompt
P -> worn mech suit, (steel metal [rust]), intricate, elegant, sharp focus, very high detailing, natural lighting, photorealistic, octane render, HDR, 8k, high contrast
N-> bad anatomy, nude, cartoon, anime, sketches, (worst quality:2), (low quality:2), (normal quality:2), lowres, normal quality, ((monochrome)), ((grayscale)) OPENPOSE
P-> tomb raider outfit, sci fi design, intricate, elegant, sharp focus, very high detailing, natural lighting, photorealistic, octane render, HDR, 8k, high contrast
N-> bad anatomy, cartoon, anime, sketches, (worst quality:2), (low quality:2), (normal quality:2), lowres, normal quality, ((monochrome)), ((grayscale)) OPENPOSE
P -> evil dark mage outfit, intricate design, elegant, blurry background, very high detailing, natural lighting, photorealistic, octane render, HDR, 8k, high contrast
N-> cartoon, anime, sketches, (worst quality:2), (low quality:2), (normal quality:2), lowres, normal quality, ((monochrome)), ((grayscale))
P -> steampunk rogue outfit, torn cloth, very high detailing, natural lighting, photorealistic, octane render, HDR, 8k, high contrast
N-> unrealistic, cartoon, bad anatomy, deformed body parts, low quality, worst quality, blur, ((monochrome))
P> (Steampunk) female evil (Knight), ((dark black)) armor, ((sexy)) glossy look, (dark lord), (intricate engraving design), sharp focus, very high detailing, natural lighting, photorealistic, octane render, HDR, 8k, high contrast
N-> ((unrealistic)), rough metal, cartoon, digital art, (bad anatomy), deformed body parts, low quality, worst quality, bad hands, blur, ((weapons))
P-> (((Rococo/Gothic/Art
nuevo/Prehistoric/Industrial/indian/Byzantine/Islamic/Expressionist/Korean/Egyptian/Mesopotamian/Tudor/Greek/Javanese/Deconstructivist/Chinese/Japanese/Renaissance/Baroque/Victorian))) outfit, very high detailin
N-> ((unrealistic)), ((cartoon)), digital art, (bad anatomy), deformed body parts, low quality, worst quality, ((bad hands)), blur
P-> (((Gothic))) outfit, masterpiece, best quality, very high intricate detailing, sharp focus, natural lighting, (((photorealistic))), octane render, HDR, 8k, high contrast
N-> ((unrealistic)), ((cartoon)), digital art, (bad anatomy), deformed body parts, low quality, worst quality, ((bad hands)), blur
P-> Harry potter style witch outfit, aged cloth, very high detailing, natural lighting, photorealistic, octane render, HDR, 8k, high contrast
N-> ((unrealistic)), ((cartoon)), digital art, (bad anatomy), deformed body parts, low quality, worst quality, blur
Fantasy, Persian, desert, sci fi, (explorer outfit), aged torn cloth, very high detailing, natural lighting, photorealistic, dehaze, octane render, HDR, 8k, (high contrast), ((exposed hands)) ((unrealistic)), ((cartoon)), digital art, (bad anatomy), deformed body parts, low quality, worst quality, ((bad hands)), blur I
masterpiece, best quality, block buster movie, sci fi theme, intricate detailing, sharp edges, very high detailing, natural lighting, photorealistic, octane render, HDR, 8k, high contrast blockbuster movie, masterpiece, best quality, star wars style Jedi armor, very high detailing, natural lighting, photorealistic, octane render, HDR, 8k, high contrast
Lovely
Thank you so much for this♥
Thank you 🥰
Is there a way to apply the same outfit to different models at once?
The thumbnail looks awesome! Do you use stable diffusion as well to create the fire and the lighting colours. Would love to see tutorial of that!
stable diffusion was used only to change outfit. the color grading was done in DaVinci Resolve : )
You can do that in SD
what if you wanna use a particular design you made yourself?
Which MacBook would you recommend to be able to work faster with stable diffusion?
How to get consistent outfit ? Like for different images, same prompt might produce different outfit
At first I was getting good results. Now most of the time I'm getting completely deformed outputs. Why is it like that? How to solve this issue.
Thanks, brilliantly explained, very clear. All the best! 🙂
Thanks
@@lights_camera_ai My pleasure! 🙂
SUBSCRIBED!!!! Geez! Thanks a lot! It worked!
Most welcome
where is the link of the prompt?
Very interesting look at how far AI has come, cool what will be possible in future.
no description?
Where your discription?
Thanks for sharing, got me thinking of karmas incredible AI generated headshots.
How to change outfit using another picture?
Only greay background is showing as result.. how to fix
please can you share your prompt. they look really great
Sure do take a screen shot
thank you :)
mans got a fire setup
haha thanks, do check out the latest videos they're hotter
where is the website ,i cant access it..
I am a new subscriber. This is really a nice tutorial. Can you please make a tutorial using control net for poses. I want to make a person with one outfit (along with background) to match a pose of another person. Kindly make a step by step method. Also how to do in batches.
Sure, let me check what I can do
Hello thank you for the nice video! As I followed along, I got curious. Is there any way to put a picture of the clothes I want to wear instead of the clothes generated by AI?
Yes you will need to specify what clothes in text
How to install stable diffusion on Mac? Any help?
Prompts please 😢
Good one!
Another way to replace clothes: Extract the pose from the same image (ControlNet Pose Full), use multi ControlNet with Pose: full without Preprocessor selected and only with pose image and model loaded + Depth (same here: without preprocessor, with model and extracted depth image loaded)+ Reference only (pick a photo with the cloth you want to use, it can be even a front view image of the cloth) Ref Only with "ControlNet is more Important" and 1.2 strength.
Now you can use any cloth you want. It doesn't give 100% consistency (no AI model gives us this), but it kinda works.
Not sure I entirely understand haha.... But thanks
@@lights_camera_ai ^^
@@Amelia_PC can u make a small tutorial ?
@@RahulR-nm3kd Sorry! Cain't ^^
I forgot to mention that you need to make the changing on the Inpainting tab. (sample of what was done: i.imgur.com/pPotMh8.jpg) If you have any specific questions about something above, feel free to ask :)
can you make tutorial for this?
Tip: you can screenshot his prompts at 3:17 and use Google Lens to scan the texts 👌
Perfect
Your pc is burning. I see it in the left upper corner😂
Awesome video! Who needs a personal shopper when you got AI amiright? Technology is wild.
description?
Thanks Bro.
Yes but how can I keep consistent characters in the same clothes across multiple images?
you can download LORA for that and the outfits shall be same
Common...it was "inpaint"!...it can change the masked area only.😂
looks great!! but why did you set the depth´s ending control step to 0.5? how did u come up with that? thanks 4 sharing
hello brother. We keep it .5 to get a balance. If we keep it 1 then the output will be almost same as the original, if we keep it 0 then the output will be crazy different than the original
@@lights_camera_ai wonderfull thanks
But it's not depth. It's Ending control steps, it has nothing to do with depth.
@@kiransurwade3576 you are right. Thats why i was wondering how ending steps affected the depth output
How do a turn a low quality photo into a high quality one? I took a photo in the dark and it's great composition, just bad quality
You can try upscaling on Stable Diffusion
i have strong piximperfect vibes here :D
Nice video bro. Can we use something else other than depth? like softedge of canny?
((anyway, at first glance I really thought your CPU is burning in the background. LOL))
haha, yea that's me going nuts with my setup
Can you send the prompt and negative prompt?
Well. I kinda' lost it as I make so many prompts everyday. sorry brother
Thanks for the video. One question, I have a model that I generated in Midjourney and I want to dress her with an SPECIFIC dress (from a brand) it has to be exactly the dress but in the AI model I created. Is that possible using Stable Difussion (I'd add a photo of the dress in HD, white background) Thank you, I really need to solve this for a client! And at the moment I am photoshopping it. I think there must be softwares to do this
Yes we need a different method
@@lights_camera_ai what would it be? Can you please explain us and make a video! Would really appreciate it 🙏🏻
Hi just subscribed to your channel. and just finish talking to Canva Support of their fashion Ai clothes change. Anyway I was just at Dall-e and your channel just came up. Are you a photographer or a designer? or a Software engineer or just a content creatoretc... There is not much Niche on how hard to get a model, money, time etc.. and make the clothes perfect, my friend always complain about these... I mean there is some channel and website, but yours is particular. Cant wait to see this channel go viral.
Thanks a lot. Fingers crossed 🤟
Can we get same outfit in every photo? For example Japan kimono style, white color+pink Obi etc. I want exactly same outfit and ornaments in all photo.
Yes yes for that we need to use LORA
@@lights_camera_ai Thank you
Hey few questions.
1 Which version r u using 1.5 or 2.1?
2 I am not able to choose style. What could be the reason? Cant generate image also. Nothing comes up after clicking generate. I am using 2.1.
3 Where to find/install control net. I dont have it in mine.
Do check youtube for control net installation
name the website?
Dude you didn't add image to controlnet, how is it gonna create the depth map?
do not add it on the controlnet. when doing image to imaging. it will fry your PC without giving any extra results
@@lights_camera_aiNo.. without image, controlnet doesn't work.
can you drop link web stable diffusion please? or make tutorial to login to the web?
Model not shown in
model is epic realism
can you share your prompts pls. they look really helpful
sure do hit me up on my Instagram and I shall share
what checkpoint did you use
Stable diffusion with Google Collab tutorial.
How do I get the link to the website you use to edit the picture I can't find it
Search "how to download stable diffusion "
did you record this video in the middle of a busy street? xD
Yes bro. I used VFX to make it look like an indoor room
What a model (MB.ckpt) do you use in video ?
it is actually a check point merge between two different models. It is not that important.
I see some smoke coming out of the PC lol
Haha sure thing
bro please share the link! ki vabe bujhbo kon software ba kon website? ei tuku guide na korle next video gulo follow kori ki kore boss! guide us properly bro
Hi bro i like your video 👍
thanks man
Wish I could change outfits that easy, be a lazy persons dream come true haha.
Bro your PC is smoking! Is it on fire? 🤣
what website is this ?
Search download stable diffusion
how do i add control net, i dont have it
You'll find lots of tutorials online to install control net. It is very easy
good
thanks brother
link for site please
Hey there this is not a site, this is something you need to install
3.55 texts are not in description please add
You can screen shot them : )
why you not add on discription its took 30 minutes to type so kndly attach file
i did not get prompt file please attach file link in the comment thank u
ping me up on insta and I'll personally share the prompt
there is smoke... something burning behind you..
Where is the website, I can't find the website link
it is not website feature, to use stable diffussion you need a pc with good graphic card and then you need to install stable diff step by step (you can find on youtube)
Where is result?
do follow the video.
where is the prompts?
Do hit me up on insta and I shall share
@@lights_camera_aiJust put the prompts in the description.
Can i dress the model with own picture ? If case i created teeshirt with photoshp.
Sure absolutely
@@lights_camera_ai please can you give me an idea how ton achieve this??? Thank uuu
@@borkisoufiane you find solution?
Give the website link there are lots of stable diffusion webs
you will need to install stable diffusion
@@lights_camera_ai any installation vedio??
@@lights_camera_aiyou ain't provide the exact link of stable diffusion
YOU MAKING PEOPLES LIVE VERY DIFFICULT
hi, can i have the text of the prompt ?
sure thing do screen shot and you can type it. Or check out the new videos there the prompts are in the description : )
@@lights_camera_ai
Why don't you just update the description of the video?
@@SodaMan18 he dont deserve a subscribe
Constructive Feedback:
1. Background noise in the video can be removed using NVIDIA RTX voice application (also works with GTX cards)
2. Edit could be short as people will see content which is short without long pauses (it is engaging)
3. In the process we could avoid the face process is important which also allows us to just record a voice over
Nice video !!
Sure sure thanks
Meanwhile Photoshop generative fill doing this in seconds..
not exactly.
can you pin the document link of command promt ?
sure do hit me up on Instagram and I shall share
@@lights_camera_ai ok thanks
Qu tarjeta grafica usas ?
Wait lemme translate
YO! Why is your CPU smoking??? 🤣
Too hot to handle hehe
@@lights_camera_ai 😂
lol nobody say anything about the PC got smoke coming out? 🤣
Nobody indeed
thx, where are the prompts?
You can pause the video and take screen shot of the prompts
Model link?
It's the stable diffusion 1.5 model. I renamed it randomly as MB
Is it for free??
yes it is
@@lights_camera_ai can i upload an image frm my pc and change it according to wish??
Bro kapde utarne ka tutorial
Easy, take off your shirt and pants
prompts ?
hey there do hit me up on insta for the prompts
@@lights_camera_ai Share ur insta
Incomplete
This can be used for evil deeds.
Haha true that
Are you doing this to make people naked?
hide your face