<p>Simple Workflow that uses the new Florence Nodes to "replace" older detection techniques that requires many specific models, (Ultralytics YOLO's)</p><p></p><p>Now you can just have one single model to detect pretty much anything with more details and accuracy, (face, eye, mouth, hairs, watermark removing, inpainting tools to change colors or shapes) Alot of usages possible,</p><p></p><p>DINO is a also a good alternative but I think Florence is better, have to do some tests</p><p></p><p>Here you can see how it works</p><div data-youtube-video><iframe width="640" height="480" allowfullscreen="true" autoplay="false" disablekbcontrols="false" enableiframeapi="false" endtime="0" ivloadpolicy="0" loop="false" modestbranding="false" origin playlist src="https://www.youtube.com/embed/ONaeAcoWwtg" start="0"></iframe></div><p>I recommend using an Inpaint model for strong changes, (above 0.4 denoise value)</p><p>You can use a classic SDXL model for "details" (under 0.4 denoise value)</p>

Files with SHA256: 5fd7ce6b8bbf700351898f9be2033e7135587eeb43a3d9803933e642d3955384