How to Do Virtual Try-On with AI (Any Product Photo)

How to do virtual try-on with AI: put any product photo on any model image, and you don't have to write a text prompt. Virtual try-on needs two things, a photo of a model and a photo of a product. The result keeps everything from the model image, the face, the body, the pose, the lighting and the background. Only the garment changes. And the garment is your product as it is in the product photo, the same cut, color, print and details. The AI puts it on her, it does not redesign it. The images decide what the result looks like, not the text. In this video I go through it inside the FASHN Agent with one model and build a full outfit on her, one piece at a time. The product photo. On-person, flat-lay and mannequin photos all work. On-person is the best because it shows how the garment fits, and I show what happens when the photo doesn't. No prompt at all. Send the two images with nothing typed and the Agent asks what you want. Tops, bottoms and a jacket. What changes and what stays the same, why a jacket goes over what she is wearing instead of replacing it, and why sending the top and bottom together beats doing them one at a time. Fixing a miss with words. The skirt came out high waisted from a flat-lay photo, so I told the Agent what I wanted and it fixed it. Then accessories, a new pose and the back view of the same look, the shots you need for a product page. Chapters 0:00 Intro 0:32 Product photos 0:57 On-person 1:18 Flat-lay 1:44 Mannequin 2:16 Top 2:32 Bottom 3:19 Top + bottom 3:45 Jacket 4:11 Accessories 4:43 Pose 5:05 Angle 5:37 Recap Questions, or something you want covered next? Leave a comment, I read every one. Try it free, you get credits when you sign up: https://fashn.ai/?utm_source=youtube