<h1 id="imagestylize-v2-or-real-to-anime-and-style-transfer-engine-(sdxl)">ImageStylize V2 | Real-to-Anime &amp; Style Transfer Engine (SDXL)</h1>Transform real-world photographic portraits into stunning, highly stylized anime illustrations without losing character identity or composition. ImageStylize V2 is a highly refined, production-ready SDXL image-to-image workflow built to push the boundaries of artistic style transfers while keeping your generations completely under control.Whether you are converting photos to vibrant cell-shaded illustrations or shifting between highly stylized artistic mediums, this workflow eliminates the usual headache of distorted faces, mutated hands, and broken compositions.<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/f287abeb-a8ba-46af-a763-ca8973b82f48/original=true/f287abeb-a8ba-46af-a763-ca8973b82f48.jpeg" /><h3 id="core-features">🚀 Core Features</h3><ul><li>IPAdapter FaceID Integration: Built-in face identity embedding loops (using FaceID PlusV2) lock down your subject's likeness. It bridges the gap between a 2D style transformation and real-world facial proportions, ensuring the character actually looks like the source image.</li><li>Dual-Layer ControlNet Stack: Combines spatial depth analysis and clean edge boundaries to ensure structural integrity. Your character's exact pose, clothing folds, and limb placement are perfectly preserved.</li><li>Smart Face &amp; Hand Auto-Detailers: Integrated Impact Pack pipelines automatically run secondary, localized inpainting passes on faces and hands. This eliminates blurred expressions and cleans up unstable AI hand geometry on the fly.</li><li>High-Res Polish Stack: Features an optimized upscaling route paired with a micro-dose image noise generator. This acts as a subtle film or cel grain, completely removing ugly digital color banding and preparing your images for crisp uploads.</li></ul><h2 id="required-models-and-directory-setup-guide">📦 Required Models &amp; Directory Setup Guide</h2>To run this workflow successfully, download the following models and place them into their respective ComfyUI directories. (Note: Most of these can also be easily searched and downloaded directly via the ComfyUI Manager's "Model Manager" tool).<h3 id="1.-base-checkpoint-and-loras">1. Base Checkpoint &amp; LoRAs</h3><ul><li>SDXL Base Checkpoint: <a target="_blank" rel="ugc" href="https://civarchive.com/models/1904102/sakura-dreams">Sakura Dreams (or your preferred SDXL Stylized Base)</a><ul><li>📂 Directory: <code>ComfyUI/models/checkpoints/</code></li></ul></li><li>IP-Adapter FaceID LoRA: <code>ip-adapter-faceid-plusv2_sdxl_lora.safetensors</code><ul><li>📂 Directory: <code>ComfyUI/models/loras/</code></li></ul></li></ul><h3 id="2.-controlnets-and-adapters">2. ControlNets &amp; Adapters</h3><ul><li>ControlNet Depth: <code>controlnet-depth-sdxl-10.safetensors</code><ul><li>📂 Directory: <code>ComfyUI/models/controlnet/</code></li></ul></li><li>ControlNet Canny/Lineart: <code>controlnet-canny-sdxl-10.safetensors</code><ul><li>📂 Directory: <code>ComfyUI/models/controlnet/</code></li></ul></li><li>IP-Adapter Model: <code>ip-adapter-faceid-plusv2_sdxl.bin</code><ul><li>📂 Directory: <code>ComfyUI/models/ipadapter/</code></li></ul></li><li>CLIP Vision Encoder: <code>CLIP-ViT-bigG-14-laion2B-39B-b160k.safetensors</code><ul><li>📂 Directory: <code>ComfyUI/models/clip_vision/</code></li></ul></li></ul><h3 id="3.-yolo-detectors-(impact-pack)">3. YOLO Detectors (Impact Pack)</h3><ul><li>YOLO Face Detector: <code>face_yolov8m.pt</code><ul><li>📂 Directory: <code>ComfyUI/models/ultralytics/bbox/</code></li></ul></li><li>YOLO Hand Detector: <code>hand_yolov8s.pt</code><ul><li>📂 Directory: <code>ComfyUI/models/ultralytics/bbox/</code></li></ul></li></ul><h3 id="4.-upscalers">4. Upscalers</h3><ul><li>Pre-Attention Face Isolation Upscaler: <code>4x-UltraSharp.pth</code><ul><li>📂 Directory: <code>ComfyUI/models/upscale_models/</code></li></ul></li><li>Final Artwork Upscaler: <code>RealESRGAN_x4plus_anime_6B.pth</code><ul><li>📂 Directory: <code>ComfyUI/models/upscale_models/</code></li></ul></li></ul>💡 Crucial Step for IP-Adapter FaceID: FaceID requires the <code>insightface</code> Python library. If you encounter a <code>ValueError</code> or compilation error on boot, run the following command in your root directory terminal to compile it against your environment:<pre><code>.\python_embeded\python.exe -m pip install insightface --no-binary insightface --no-build-isolation --force-reinstall</code></pre><h3 id="recommended-starter-settings-(real-anime)">🎛️ Recommended Starter Settings (Real ➔ Anime)</h3>To get the absolute most out of this styler right out of the box, load up an SDXL anime/stylized checkpoint and dial in these baselines:<ul><li>KSampler Denoise: <code>0.60 - 0.70</code> (The sweet spot for giving the model enough power to redraw textures without losing the original photo layout)</li><li>CFG Scale: <code>3.5 - 5.0</code> (Keeps colors vibrant without clipping lines or burning the image)</li><li>Sampler/Scheduler: <code>dpmpp_2m</code> / <code>karras</code></li><li>ControlNet Weights: Depth (<code>0.65</code> strength, ends at <code>0.60</code>) | Edge/Lineart (<code>0.35</code> strength, ends at <code>0.75</code>)</li><li>Face / Hand Detailer Denoise: Face (<code>0.50</code>) | Hand (<code>0.40</code>)</li></ul><h3 id="required-custom-nodes">📦 Required Custom Nodes</h3>Make sure you have updated these suites via the ComfyUI Manager before running:<ol><li><code>ComfyUI-Easy-Use</code> (For the optimized modular loader loops)</li><li><code>ComfyUI-Impact-Pack</code> (For the segmentation, detection, and detailer loops)</li><li><code>ComfyUI-ControlNet-Auxiliary-Preprocessors</code> (For clean depth and boundary tracking)</li><li><code>ComfyUI_IPAdapter_plus</code> (For facial vector integration)</li><li><code>ComfyUI-KJNodes</code> (For final noise/grain polish)</li></ol>

Files with SHA256: 2d71a8db8ea409639f2874da01a810924f0aadc73e5d58e5dda1dbc8612cd931