AI for Sewists: Creating Your Own AI Model for Garment Visualization

1. Course Intro: Greetings. I'm Vanessa S of S by VanessaS and welcome to my course, AI for Sovis creating your own AI model for garment visualization. In this course, I'm going to show you how to create your own AI model in your likeness or similarities to your likeness. And these can be used for your sewing projects so that you can visualize the garment that you want to make with the fabric that you want to use and how it looks on your body. In this course, we're going to use Gemini and Canva to create the visual models, as well as to create the final product. At the end, there's going to be a class project, which I will discuss more in that lesson. So if you're ready, I'm ready, I will see you in the first lesson. 2. Lesson: Creating AI Model: Welcome back. So when creating a model of your likeness or similarities to your likeness, there's a few things that you need to do for your prompt. Now, understand that there's two ways that you can do this. One is you can create a text prompt or you can do a photo upload. The text prompt, you will need to add a description. This description normally has the race, the skin tone, the hair type, hair color glasses, body type, things like that. And it will provide you with more privacy because there's no photo required. And the result is you'll have a likeness, but it will not be an exact replica or match of you. However, it will be similar. So if you tell the prompt that you want a Black woman with locks and glasses, more than likely, that's what you're going to get. Now, she may not look like you, but she will be a Black woman with locks or glasses. The photo upload, on the other hand, is a little bit different. If you upload a photo of yourself directly into Gemini or Canva, you're going to get something closer to your likeness and a little bit more accurate results. But you have to consider your comfort level with sharing your image with AI, regardless of the platform. So a typical prompt would sound like this. And what I'm going to do is I'm going to show you the prompt results for both Gemini and for Canva, we're going to use the same exact prompts for both the text prompt description and the photo upload. And I'm going to show you the results for both, and then we're going to compare them. Okay, so the text description prompt is create an image because you always want to start with telling AI what you wanted to do. So in this case, it is create an image, a full body studio portrait of a Black woman with brown skin tones and natural locks standing in a neutral pose against a solid light gray background, even flat lighting, wearing simple form fitting white leggings and a white tank top. Now, understand why this is important because the goal is to take your virtual model and use your model for future projects, regardless of what the fabric, the print, regardless of what it is, you can use that one model that you're creating for different projects. So if you have her neutrally dressed, then the clothing is not going to be an issue. Okay. And so for the photo upload prompt, you're tweaking it just a little bit to say, you want a full body studio portrait, but you want it to use the attached image, okay? Alright, so here are the results. So the text description provided me with her. Now, the photo upload, the photo that I uploaded, is this young lady right here, and notice that the ness is the same, but it did provide me with the garment, the background, everything else, as I requested. So the text description gave me a very general looking young lady. And then the photo upload gave me an exact replica of the young lady in the photo. With Canva, you do the exact same thing. You use the exact same prompt. But you're going to find that there's some differences. One is, Canva is going to give you more iterations, whereas Gemini will give you one image unless you ask it specifically for multiple images. Gemini is going to give you one image. Canva is going to give you four. Now, these are the iterations that I received the results from the same exact prompts. So text description. This is the young lady that I received. And notice not only do I have four different images, it's the same young lady, but it's different poses. For the photo upload, I use the same exact photo, and this is what was returned to me. So notice that Canva does not provide the same exact photo image, even though I uploaded it and stated the same thing. Let's do an overall comparison. For the text description, Gemini and Canva both gave us what we asked for. Canva, however, was a little bit more creative. For the photo upload, Gemini and Canva both gave us what we asked for, but Gemini was more on point as to the actual description and the request. It's a preference. You can decide which is better or which is worse. The whole point here is that both platforms can help you when it comes to your sewing projects and using models that actually look like you so that you can see how your garments are going to look and feel. Now, we have one more step when it comes to using your model and your likeness or similarities to your likeness for your sewing projects. And here's my favorite part. One is you take the model, whichever you choose, the one that you like the most. You take your model, you add a pattern, an image of the pattern or either a picture of the actual garment that you're wanting to make and then you can add a swatch or a picture of the fabric that you want to use. And the prompt is really simple. Create an image of the model wearing the attached dress in the attached fabric with a natural pose. That's all you're asking. What you are trying to do here is get a visual of what this garment is going to look like in the fabric that you want and then what it's going to look like on you. That's the whole visualization, it's helping you as a sovis because most of what sewing is is visual outside of the actual construction of the garments. It's visual. So now you have a way to figure out what this garment is gonna look like on you before you have to do anything else. And the results there you have it. So we have our model, which we created in our likeness or similarity to our likeness, the pattern that we're interested in making and the fabric that we would like to make it in, and this is what it would look like for you. So the workflow for this is really simple. It's really simple. If you follow the steps, it'll go by quickly. Create a prompt that describes the visual features of your model, hair, texture, skin complexion, glasses, no glasses, whether you want jewelry, whether you're a little pudgy, whether you're a little thin, you can decide. Use the prompt and Gemini and Canva. You can use either one or both. And then if you need to tweak it just tweak it, revise the prompt and regenerate the model, and then download your image to your photos, create a prompt to use for the sewing garment of interest, add a swatch of the fabric, the type of garment. If you have an image, attach the image and attach the image of your model and submit. That will render you a model in your chosen garment for your visualization and to help you with your sewing project. 3. Class Project: Let's talk about your class project. For your class project, you're going to create your own AI model in your likeness or similarities to your likeness. And I keep saying similarities to your likeness because if you're using text to the text description for your prompt, that's similar to your likeness. But if you're using your photo, that is your likeness. So it's either similarities to your likeness or your likeness, depending on which process you choose. So you're going to choose your tool, Gemini or Canva. You're going to choose your approach, meaning text description or photo upload. You're going to create your prompt, and then you're going to generate your visual model. The second part is create a prompt to use for the garment itself. Okay? Whether it's going to be a skirt or jacket, pants, whatever the garment is, create a prompt to use for the sewing garment of interest. Add a swatch of the fabric. Remember, you can take a picture, you can go online and get a swatch, tile, however you want to do. The type of garment, if you have an image, attach the image and attach the image of your model and submit. The image should render your model in the fabric choice and garment choice of your choosing. Upload the image to the class project folder, and that's it. So I can't wait to see what you create.