Runware/adetailer
13.8k
1---2license: apache-2.03library_name: ultralytics4datasets:5- wider_face6- skytnt/anime-segmentation7tags:8- pytorch9---10 11# YOLOv8 Detection Model12 13## Datasets14 15### Face16 17- [Anime Face CreateML](https://universe.roboflow.com/my-workspace-mph8o/anime-face-createml)18- [xml2txt](https://universe.roboflow.com/0oooooo0/xml2txt-njqx1)19- [AN](https://universe.roboflow.com/sed-b8vkf/an-lfg5i)20- [wider face](http://shuoyang1213.me/WIDERFACE/index.html)21 22### Hand23 24- [AnHDet](https://universe.roboflow.com/1-yshhi/anhdet)25- [hand-detection-fuao9](https://universe.roboflow.com/catwithawand/hand-detection-fuao9)26 27### Person28 29- [coco2017](https://cocodataset.org/#home) (only person)30- [AniSeg](https://github.com/jerryli27/AniSeg)31- [skytnt/anime-segmentation](https://huggingface.co/datasets/skytnt/anime-segmentation)32 33### deepfashion234 35- [deepfashion2](https://github.com/switchablenorms/DeepFashion2)36 37| id | label |38| --- | --------------------- |39| 0 | short_sleeved_shirt |40| 1 | long_sleeved_shirt |41| 2 | short_sleeved_outwear |42| 3 | long_sleeved_outwear |43| 4 | vest |44| 5 | sling |45| 6 | shorts |46| 7 | trousers |47| 8 | skirt |48| 9 | short_sleeved_dress |49| 10 | long_sleeved_dress |50| 11 | vest_dress |51| 12 | sling_dress |52 53## Info54 55| Model | Target | mAP 50 | mAP 50-95 |56| --------------------------- | --------------------- | ----------------------------- | ----------------------------- |57| face_yolov8n.pt | 2D / realistic face | 0.660 | 0.366 |58| face_yolov8n_v2.pt | 2D / realistic face | 0.669 | 0.372 |59| face_yolov8s.pt | 2D / realistic face | 0.713 | 0.404 |60| face_yolov8m.pt | 2D / realistic face | 0.737 | 0.424 |61| face_yolov9c.pt | 2D / realistic face | 0.748 | 0.433 |62| hand_yolov8n.pt | 2D / realistic hand | 0.767 | 0.505 |63| hand_yolov8s.pt | 2D / realistic hand | 0.794 | 0.527 |64| hand_yolov9c.pt | 2D / realistic hand | 0.810 | 0.550 |65| person_yolov8n-seg.pt | 2D / realistic person | 0.782 (bbox)<br/>0.761 (mask) | 0.555 (bbox)<br/>0.460 (mask) |66| person_yolov8s-seg.pt | 2D / realistic person | 0.824 (bbox)<br/>0.809 (mask) | 0.605 (bbox)<br/>0.508 (mask) |67| person_yolov8m-seg.pt | 2D / realistic person | 0.849 (bbox)<br/>0.831 (mask) | 0.636 (bbox)<br/>0.533 (mask) |68| deepfashion2_yolov8s-seg.pt | realistic clothes | 0.849 (bbox)<br/>0.840 (mask) | 0.763 (bbox)<br/>0.675 (mask) |69 70## Usage71 72```python73from huggingface_hub import hf_hub_download74from ultralytics import YOLO75 76path = hf_hub_download("Bingsu/adetailer", "face_yolov8n.pt")77model = YOLO(path)78```79 80```python81import cv282from PIL import Image83 84img = "https://farm5.staticflickr.com/4139/4887614566_6b57ec4422_z.jpg"85output = model(img)86pred = output[0].plot()87pred = cv2.cvtColor(pred, cv2.COLOR_BGR2RGB)88pred = Image.fromarray(pred)89pred90```91 9293 94 95## Unsafe files96 9798 99Since `getattr` is classified as a dangerous pickle function, any segmentation model that uses it is classified as unsafe.100 101All models were created and saved using the official [ultralytics](https://github.com/ultralytics/ultralytics) library, so it's okay to use files downloaded from a trusted source.102 103See also: https://huggingface.co/docs/hub/security-pickle104