About the prediction of yolo boundary box prediction

[JJaassoonn]

Dear Mentor,

I found that in yolo_head() function, the prediction of bounding box on width and height is unscaled from log-scale to linear scale, and box_wh is unscaled relative to anchor boxes.

1. box_wh = K.exp(feats[…, 2:4])
2. box_wh = box_wh * anchors_tensor

Does it mean the width and height of a bounding box is scaled with respect to width and height of an anchor box , and applying log scaling?

log(width of bounding box / width of anchor box)
log(height of bounding box / height of anchor box)

Then, the log-scaled width and height are annotated in the truth label in the training set?

Thank you.

[ai_curious]

Applying YOLO anchor boxes

Applying YOLO Anchor Boxes A previous thread provides an introduction to extracting from labelled ground truth data a small set, on the order of 3 <= k <= 10 , of shapes for what is referred to in the YOLO 9000 paper as dimension clusters or priors, but in this class as anchor boxes . NOTE: Just to make things more confusing, the YOLO 9000 paper uses anchor boxes to refer instead to the hand-picked shapes described in the Faster R-CNN paper, not the ones generated by K-Means. The YOLO 9000 pap…

Maybe that will help?

Let us know