Generative image app alternatives - Midjourney Tutorial

- [Instructor] In this video, we're going to take a look at some other text-to-image based alternatives besides Midjourney. We're going to start off by looking at DreamStudio, which is still in beta, beta.dreamstudio.ai, and it actually taps into a model called Stable Diffusion. Like other services that we saw earlier, you basically purchase a certain series of credits, and with those credits you can generate imagery. With a free account, you get a certain amount of credits that you can start with. In looking at DreamStudio, on the left-hand side, you can first of all choose the model that you'd like to work with. So, right now we can see the various Stable Diffusion models that we can generate images with, including a beta for Stable Diffusion XL. And next, we can choose a style of art that we would like to base the prompt off of. You can see some other ones loading up here. I'm going to choose Low poly in this instance. And then, we can start to write the prompt of what we're looking to create. In this case, I'm going to type in, "A field of playful cartoon animals dancing to music." Now, in terms of a negative prompt, very similar to the --no parameter that we added during one Midjourney text prompt, we can set that up here of what we might not want to see. So, let's go with lions, no tigers, and, in this case, bears. (laughs) In some cases, we could also upload an image to use along with our text prompts. And depending on what model you're using, you can change the aspect ratio of the image you'd like, anywhere from a 9x16 to 1x1, and then your standard 16x9 based images. Next, you have the option of choosing how many images you would like DreamStudio to produce. In this case, I'll do the standard form, and then we're off to the races where we can Dream. So, click on Dream, we'll see it get to work. And our four images generated in that low poly style. Very similar to Midjourney, we have a couple options here where once we find an image that we like, I really kind of like this one here, we can create some variations of it, and there we can see the end results. We can download the image if required. Another generative text-to-image based app that we can take a look at is Adobe Firefly, which also at the time of this recording is in beta. There's a few different things that are within Firefly for generative art creation, but let's take a look at the classic text to image. An example of images that people have come up with. And if we'd like to, we can try a prompt and then it will come up with four variations. We can rate on if we like what it's come up with based on the prompt. And in addition, of course, we can make changes to this. So, right now it's says, "A cute, fluffy blue ball animal running through a magical forest, trying to catch a flying firefly." Let's change this to, "A cute green ball animal," and also pay attention to the styles that are here: Photo, Misty, Beautiful. In this case, I'm going to get rid of Photo. I like Misty and Beautiful here. And under this Styles section here at the side, I want to just choose Techniques. You can see various techniques that we can choose from. I'm interested in an Acrylic painting in this case. I'm going to Generate, Firefly gets to work, and we can see here the images that it created. I'm going to have it Refresh just to see if I get something else differently here with my Acrylic painting style added. So, here's the end results after the Refresh generation. And similar to what we've seen, we have options here to Download. And we can click on this to Show similar results to this image, and it's going to generate a few additional frames in that style. When we find what we like, simply download that. Now, this page really shows what Adobe is all about in being transparent with their AI. It will include information that this image was generated with AI when you share it with other people. And in addition to that, what makes Firefly unique is Adobe Firefly is trained its model using primarily images from its Adobe Stock collection, which it has the absolute rights to. Let's choose Continue, and then Allow our contents to be downloaded. Or we can see the image, and at the time of this beta that this is not for any form of commercial use. So, there's a few other generative text-to-image AI apps that you should become familiar with. But DreamStudio, which is tapping into Stable Diffusion, as well as Adobe's Firefly currently in beta.