<h1 id="anthro-hands-detailer-or-segmentation-model-(yolo26s-seg)">Anthro Hands Detailer | Segmentation Model (YOLO26s-seg)</h1>A single-class hand segmentation model for ADetailer, trained specifically on AI-generated anthro art - pawpads, claws and all the poses stock human-hand detectors choke on. Drop it into your A1111 / Forge / SD.Next or ComfyUI install, point ADetailer at it, and let it find and repaint the hands in your anthro generations.This one is a instance segmentation model: ADetailer gets an actual per-hand mask instead of a rectangle stretched over the bounding box, so inpainting touches the hand (and just enough of the wrist) instead of a chunk of the background.<h2 id="quick-facts">Quick facts</h2><ul><li>Architecture: YOLO26s-seg (end-to-end, NMS-free)</li><li>Classes: <code>hand</code></li><li>Native resolution: 1024 (letterboxed, gray padding)</li><li>Training data: ~5,500 AI-generated anthro art images, ~9,000 hand masks</li><li>Size / speed: ~10M params, ~22 MB</li></ul><h2 id="validation-results-(v1.0)">Validation results (v1.0)</h2>Measured on a held-out validation set of 566 images / 993 hand instances (best checkpoint):<ul><li>Mask: Precision 0.950, Recall 0.937, mAP@50 0.968, mAP@50–95 0.811</li><li>Box: Precision 0.948, Recall 0.934, mAP@50 0.967, mAP@50–95 0.821</li></ul>In practice this means: it rarely invents hands that aren't there (high precision), and it finds hands down to fairly small sizes.<h2 id="installation">Installation</h2>Automatic1111 / Forge / SD.Next1. Download the <code>.pt</code> file.2. Copy it to <code>stable-diffusion-webui/models/adetailer/</code>3. Restart the WebUI (full restart if it was already open).4. Select the model in the ADetailer tab.&gt; Tip: in Settings → ADetailer you can raise the number of ADetailer passes (up to 10) if you want to run a hand pass together with face/eyes detailers.ComfyUI (Impact Pack)1. Copy the <code>.pt</code> file to <code>ComfyUI/models/ultralytics/segm/</code>2. Select it as the SEGM model in the UltralyticsDetectorProvider node.RequirementsThe model uses the YOLO26 architecture and needs <code>ultralytics &gt;= 8.4</code> (January 2026). If ADetailer throws an error on load, upgrade ultralytics inside your WebUI's venv:<pre><code>venv\Scripts\python.exe -m pip install -U ultralytics</code></pre>(portable A1111: run from the WebUI root; Forge/SD.Next — same idea, upgrade in the environment the WebUI runs from).<h2 id="recommended-adetailer-settings">Recommended ADetailer settings</h2>Starting point that worked well for me:<ul><li>Confidence threshold: 0.25–0.40 (drop to ~0.15–0.20 for small or partially hidden paws)</li><li>Inpaint denoising strength: 0.30–0.40 (up to ~0.5 if the hand is badly melted)</li><li>Use the ADetailer TAB to check detections before committing a full render.</li></ul><h2 id="notes">Notes</h2>The training set is anthro art, so anthro hands are the strong suit; human hands are outside the training distribution and may be detected inconsistently.If it saves you a reroll, a review and a posted result are the best thanks. Feedback is welcome and directly feeds the next version.

Files with SHA256: 1f5e3055ce2f40d1ef139136f63b0d6252b9f85733fcb40887d484e9fbb394fc