ComfyUI Workflow<ul><li><a target="_blank" rel="ugc" href="https://civarchive.com/articles/4916/how-to">How to install custom nodes, facelib, symlinks...</a></li><li><a target="_blank" rel="ugc" href="https://civarchive.com/articles/4642/links-resources">Resource links</a></li></ul><ul><li>you can quickly navigate in the workflow with the num keys 1-5</li><li>reduced to a minimum (compared to the older flows) for artistic but also realistic vid2vid creations.</li><li><a target="_blank" rel="ugc" href="https://youtu.be/0QjlMU00NJ8?si=MpV40GNTZ11mDEdJ">excellent beginner tutorial</a></li></ul>

Vid2Vid (Advanced Version) using AnimateDiff, ipadapter, controlnets, replace background, face swap/restore, some filters+color match, img2img

based on the scarface workflow v0.9 with some updates. squeezed into the layout of v2.Changes v2 -&gt; v3<ul><li>rework of background replacement (interpolates frames of too short videos, color match of bg/fg, fg/bg, switch of both as side effect)</li><li>added scale by factor (image settings)</li><li>added latents group to either work with the input images (vae encode) or empty latents</li><li>reworked mask area</li><li>added aio blend (combines lineart+hed in its current state, can be used with lineart or softedge models)</li><li>added wd14 auto prompt</li><li>rework of animatediff group</li><li>added a cleanup to free memory</li><li>removed hires-fix, added a 2nd sampler (upscale)</li><li>reworked/merged filter groups</li></ul>

Files with SHA256: fc45672682aed6228e2fb59586ab557188dcca3a10795416b7b66f0cd74010e8