Open rrrtype opened 5 years ago
I train tiny yolo v3 in a very limited situation.
mAP was 99.6%, loss avg was 0.23. Is it possible for the detection accuracy to be 100% with the image used for new training?
with the image used for new training?
What do you mean?
I train tiny yolo v3 in a very limited situation.
mAP was 99.6%, loss avg was 0.23. Is it possible for the detection accuracy to be 100% with the image used for new training?