Leonardo AI Image-to-Image Tutorial

Leonardo's AI image generator allows you not only to use a text prompt but to also use an existing Source image along with your text prompt to generate a new image let's look at how it works I'm on the Leonardo homepage and I need to get to the image generation page which I can do by clicking the big pink button over on the right that says create new image I can also come over here and click this big image generation button that's toward the left or image generation under AI tools on the left menu any one of those will work once we're on our AI image generation page we're just going to click on this image guidance tab right here once we've clicked that image guidance tab we need to get an image to be our guide we can drop this down right here and see our recent images or we can click this sort of upload button and it'll give us a lot more options this page will pop up and we can either upload an image from our computer pick another image that we've already uploaded we can go to Our Generations things that we've already created we can go to the Community Feed and use an image that someone else has generated or we can go to our follower feed if we had anything in there we're going to go back to the Community Feed I'm going to grab this image right here now if you're not on a paid plan you have one image to use here in image guidance and the only option you have in image guidance is image to image if we drop down this menu you'll see that Leonardo has all kinds of different ways that you can use IM guidance but all of them except image to image are just in the premium plans I get a little warning over here this yellow thing that says that my Dimensions don't match and that's not a problem I can just come right over here to advance controls and since this is saying it's 4 to three we'll just drop this down and I will pick four to three and now my little warning has gone away now we can type a prompt and we need to tell Leonardo how much we want it to weigh this image in the generation up here in the prompt I think we'll say something like we'll say a cabin on a Hillside and I'm going to leave that strength AT3 for the moment and let's say generate now to see our generation we're going to need to come back over to the generation history tab fortunately it'll leave this little green box that says on next to the image guidance Tab and that's telling us that we're using a source image with image guidance that's important because if you're finished with this image that you're working on and say you're were trying to create an image of a car or a person it would be referencing this image still that we used as a source and things probably wouldn't be working the way you expected all right so let's see what we ended up with it kept much of the original image as far as we have the mountains in the background we have this green Hillside we have those beautiful clouds in the blue sky and that was kind of what we were looking for this one took a little bit of a different angle but it definitely got us our cabin there now we had that set at a strength of. 3 so let's see what happens if we move that up to say .9 I'm still going to leave the prompt as a cabin on a hillside we'll hit generate on that and now we really don't have a cabin on our Hillside and why might that be that's because by setting this at 0. n the highest possible strength we can we told it to adhere extremely closely to the original image that we provided the source image this fell right here and as you can tell by what we got back we got back almost that same image so this strength slider has a huge impact on how your image to image Generations come out now I've been using the Leonardo diffusion XL model you can use other models we're going to drop that down and we're going to pick 3D animation style I'm going to go ahead and turn off Alchemy because I'm now down to 50 tokens for today and I would like to conserve them a bit so let's turn that off and let's go ahead and drag this strength back down to about 30 and hit generate I see I ended up with a little warning over here that I should have changed my image Dimensions before I started and I did not do that but that's okay you can tell it was not the end of the world so let's take a look at what happened for our 3D style I think that came out pretty darn good these would look great for or an animated video let's switch up our source image a little bit here this time I'm going to go back to the Community Feed okay we're going to grab this guy right here there's certainly a style to this image so we'll see how much of that we can get to carry through I'm going to say a young man with green hair and an earring because I don't see an earring in that picture so I want to see if it'll add it so I've got this little warning over here and I know what it's telling me it's saying it wants me to go to 640 by 8:32 because there's a star right there but if I hit that then I end up with this little Caution that says hey this image size is off so I come over here and I change it to 2 three which is what it's telling me at once and I end up back there so I could go in a circle all day we'll skip all that and I'm just going to generate let's go over here and see what it comes up with I forgot we left our 3D animation Style on so this is what it came up with it it definitely added our earring I don't know what those earrings are maybe those are lima beans or something it carried through in that green theme that was going on and it added an earring to every single one of these images now let's switch it back to a different model I'm going to go with Kino this time generate this is still referring to that original image this fella right here just because we generated 3D images it didn't pick that up that's our original image and and this is what it came up with and the only uh thing that I changed about this was I told it with an earring I see the earring there and that definitely fits my prompt of a young man but it also very much fits that style of the image that we started with as our source image there are so many things that you can do with image to image depending on how you tweak The Prompt which model you use and how much strength you give your Source image possibilities are endless now if you have a paid subscription for Leonardo you have all of these options available to you for image guidance not just image to image the best thing to do here let me show these to you on this screen so this is depth to image and the left here is the original image and then you see some results that were created by using the depth to image type in the image to image generator oo that was a mouthful if you want to create that depth and dimension in your images the depth to image is a pretty handy way to do it from an existing image that you already have you also have Edge to image and it has identified in the original sort of the outline of what this object is and then based on the prompt that was entered is keeping that outline of that object and changing what's within the shape and it does the same thing in result number two here now it does affect the background as you can tell by the colors we've got on the wall here you like finish finished coloring books but you don't like to go through the hassle line Art's your friend you can take just a line art drawing and suddenly make it a colorful Masterpiece with the right shading and highlights and all that good stuff normal map is Handy for 3D Graphics so you start with a 2D image you add textures and depth and you end up with something that really does appear well realistic youve even got pattern to image you can start with a pattern and then based on the prop that you type in you can end up with results that incorporate or overlay that pattern you also have pose to image and this is where you start with a pose so you've got this guy in this particular pose and then you use your prompt to take a man standing in this particular pose and create whatever you want it's not even the same face or person there it's just the same pose and you've changed everything else about this image QR to image you can take the plain old black and white QR code and you can dress that thing up and keep it still completely functional but have it look a lot cooler sketch to image let you take a very simple sketch as you can see here this thing that sort of looks like a cat and turn it into something that really looks like a cat and last but not least we have text image input so you start with just plain old text there on the screen I mean black and white how simple could it get you add your prompt you can turn it into something like this if you're not using Leonardo AI yet and you'd like to have some fun generating these images and videos and whatnot and save yourself some time finding material for your content creation efforts there's a link in the description it is an affiliate link so if you end up going with a Premium plan at some point I may receive a small commission the free version might be all you need or as you get in and start playing around you might want to explore more with other features like I did and want to keep creating more and decide to upgrade and I very much appreciate when anyone uses my affiliate link thank you sincerely

Channel: Excelerator

Views: 1,489

Keywords: leonardo ai, leonardo ai tutorial, leonardo ai image generator, leonardo ai features, how to use leonardo ai tutorial, leonardo ai training, leonardo ai tips, leonardoai, leonardo image to image, leonardo ai image to image, leonardo image guidance, leonardo ai image guidance, leonardo ai image guidance tutorial, how to use leonardo ai, leonardo ai tutorial for beginners, how to use leonardo ai image to image

Id: i6q5DLUYUCs

Length: 9min 40sec (580 seconds)

Published: Wed May 15 2024