Closed JaywongWang closed 5 years ago
Hi, I calculated mAP (area under the PR-curve) using mAP = np.trapz(DSFD[:, 0], DSFD[:, 1]), but got 0.479 instead of the reported 0.516 from the paper. I am not sure why.
Thank you for informing, the difference is caused by a bug in the code, we will update the arxiv version soon.
Hi, I calculated mAP (area under the PR-curve) using mAP = np.trapz(DSFD[:, 0], DSFD[:, 1]), but got 0.479 instead of the reported 0.516 from the paper.
I am not sure why.