deepghs/yolo-face

ONNX model exported from akanametov/yolo-face.

Use this with dghs-realutils

pip install dghs-realutils

from realutils.detect import detect_real_faces

print(detect_real_faces('yolo/solo.jpg'))
# [((168, 79, 245, 199), 'face', 0.7996422052383423)]
print(detect_real_faces('yolo/2girls.jpg'))
# [((721, 152, 1082, 726), 'face', 0.8811314702033997), ((158, 263, 509, 714), 'face', 0.8745490908622742)]
print(detect_real_faces('yolo/3+cosplay.jpg'))
# [((351, 228, 410, 302), 'face', 0.8392542600631714), ((384, 63, 427, 116), 'face', 0.8173024654388428), ((195, 109, 246, 161), 'face', 0.8126493692398071)]
print(detect_real_faces('yolo/multiple.jpg'))
# [((1074, 732, 1258, 987), 'face', 0.8792377710342407), ((1378, 536, 1541, 716), 'face', 0.8607611656188965), ((554, 295, 759, 557), 'face', 0.8541485071182251), ((897, 315, 1068, 520), 'face', 0.8539882898330688), ((1194, 230, 1329, 403), 'face', 0.8324605226516724)]

For more information, see documentation of realutils.

Models Available

Model	Type	FLOPS	Params	precision(B)	recall(B)	mAP50(B)	mAP50-95(B)	Labels
yolov11s-face	yolo	21.5G	9.43M	0.864	0.626	0.713	0.392	face
yolov11m-face	yolo	68.2G	20.1M	0.876	0.656	0.742	0.414	face
yolov11l-face	yolo	87.3G	25.3M	0.872	0.664	0.746	0.419	face
yolov10s-face	yolo	24.8G	8.07M	0.858	0.621	0.708	0.391	face
yolov10n-face	yolo	8.39G	2.71M	0.835	0.576	0.658	0.357	face
yolov10m-face	yolo	64G	16.5M	0.865	0.639	0.728	0.405	face
yolov10l-face	yolo	127G	25.8M	0.867	0.65	0.739	0.416	face
yolov8n-face	yolo	8.19G	3.01M	N/A	N/A	N/A	N/A	face
yolov8m-face	yolo	79.1G	25.9M	N/A	N/A	N/A	N/A	face
yolov8l-face	yolo	165G	43.6M	N/A	N/A	N/A	N/A	face
yolov6n-face	yolo	11.9G	4.24M	0.865	0.631	0.714	0.39	face
yolov6m-face	yolo	162G	52M	0.881	0.683	0.763	0.425	face
yolov11n-face	yolo	6.44G	2.59M	0.852	0.581	0.665	0.359	face