Question 6
Suppose you run non-max suppression on the predicted boxes below. The parameters you use for non-max suppression are that boxes with probability ≤ 0.4 are discarded, and the IoU threshold for deciding if two boxes overlap is 0.5. How many boxes will remain after non-max suppression?
My choice:
motorcycle(0.58)
pedestrian(0.98)
car(0.73)
tree(0.74)
I think there are 4 boxes will remain after non-max suppression, but the answer is wrong. Anyone can help me with this quiz,thanks!
Hmm, I think the issue here is probably with tree:
- Suppress Overlapping Boxes:
- If the IoU between a remaining box and the selected box is greater than a predefined threshold (e.g., 0.5), the remaining box is suppressed (discarded). This is because it’s considered redundant.
As I can see from the figure the IOU between the small and the big box of the tree is IOU<0.5. Hence those 2 boxes will remain there separately there.
1 Like
Hi gent.spah:
Thanks for your reply, I got that, thank you very much!
1 Like