search

Cartucho / mAP

mean Average Precision - This code evaluates the performance of your neural net for object recognition.
Apache License 2.0
2.93k stars 910 forks source link

map #147

Open alicera opened 3 years ago

alicera commented 3 years ago

If running the main.py, the mAP IOU is 0.5 or 0.5:0.75 ?

chowravc commented 3 years ago

I think it's 0.5

Refer to lines 12, 572 and 577 of main.py

alicera commented 3 years ago

Can it set the IOU 0.5:0.75 to eval？

chowravc commented 3 years ago

Try changing line 577 (main.py) from: if ovmax >= min_overlap: to: if ovmax >= min_overlap and ovmax <= 0.75:

I haven't tried it but I think that should work.

alicera commented 3 years ago

Thanks for your advance. But the it should mean with IOU 0.5:0.75.

chowravc commented 3 years ago

min_overlap is already defined to be 0.5 at the start of main.py. I think the variable name is different, but it gets passed to min_overlap.

alicera commented 3 years ago

Try changing line 577 (main.py) from: if ovmax >= min_overlap: to: if ovmax >= min_overlap and ovmax <= 0.75:

I haven't tried it but I think that should work.

if set ovmax >= min_overlap and ovmax <= 0.75 So it show the ovmax 0.9 is fail. But it is good detect when ovmax = 0.9.