Nodes/ComfyUI-yolov5-face/Yolov5 Face Detect
ComfyUI Node

Yolov5 Face Detect

A ComfyUI node in ComfyUI-yolov5-face with 5 inputs and 2 outputs.

By JiSenHua·Created about a year ago·Updated 2 months ago· 3
Yolov5 Face Detect
  • image
  • marked_image
  • face_crop
weights_nameyolov5s-face.pt
expand_ratio1.0
offset_x0.00
offset_y0.00
CategoryComfyUI-yolov5-face

Inputs (5)

NameTypeDefaultDescription
imageIMAGEInput image to detect faces from.
weights_nameCOMBOyolov5s-face.ptSelect the YOLOv5 face detection model to use.
expand_ratioFLOAT1.01–3Expansion ratio for face cropping. 1.0 = original face box, 2.0 = double the size.
offset_xFLOAT0.00-1–1Horizontal offset for face cropping. Positive = right, negative = left. Range is relative to face size.
offset_yFLOAT0.00-1–1Vertical offset for face cropping. Positive = down, negative = up. Range is relative to face size.

Outputs (2)

NameTypeDescription
marked_imageIMAGE
face_cropIMAGE