Thanks for your work!
But I got a problem using my own dataset!
It said zero-size array to reduction operation maximum which has no identity that is caused by a empty array in this line bg_score = np.power(1 - np.max(v, axis=0, keepdims=True), alpha) in myTool.py, and I delete the cam_label thresh code if cam_label[i] > 1e-5:, it runs successfully but I got a negative dloss, I don't know whether it is correct?
Thanks for your work! But I got a problem using my own dataset! It said
zero-size array to reduction operation maximum which has no identity
that is caused by a empty array in this linebg_score = np.power(1 - np.max(v, axis=0, keepdims=True), alpha)
inmyTool.py
, and I delete the cam_label thresh codeif cam_label[i] > 1e-5:
, it runs successfully but I got a negative dloss, I don't know whether it is correct?