Nodes/ComfyUI-YOLOE26/YOLOE-26 Detection Metadata
ComfyUI Node

YOLOE-26 Detection Metadata

Run prompt segmentation and return structured JSON metadata (boxes, scores, class names, mask areas) without producing image outputs.

By Rinne414·Created 5 months ago·Updated about a month ago· 11
YOLOE-26 Detection Metadata
  • model
  • image
  • metadata_json
  • detection_count
promptperson
conf0.10
iou0.70
max_det300
mask_threshold0.50
imgsz640
CategoryYOLOE26

Inputs (8)

NameTypeDefaultDescription
modelYOLOE_MODEL
imageIMAGE
promptSTRINGpersonText prompt for open-vocabulary segmentation. Separate multiple classes with commas.
confoptFLOAT0.100–1Detection confidence threshold.
iouoptFLOAT0.700–1IoU threshold used by Ultralytics inference.
max_detoptINT3001–1000Maximum number of detections returned per image.
mask_thresholdoptFLOAT0.500–1Threshold used when measuring instance mask area. Recent Ultralytics releases already return binary masks, in which case values below 1.0 have no additional effect and exactly 1.0 empties every mask.
imgszoptINT64064–2048Inference image size.

Outputs (2)

NameTypeDescription
metadata_jsonSTRING
detection_countINT