Zhang-HM

The results are too low!!!

15

I re-run the code you provided, the mAp of Pascal VOC2007 is 42.2%, but 51.5 in your paper. And in the first phase, the shot is set to 200? 

About using three times of k-shot data in the second stage of training?

1

In the second stage of training, three times of k-shot data were used？ if phase == 2: self.shots = shots * 3

Excuse me, have you tried using fast_rcnn without the RPN module. I have tried, but the result is not ideal. Is it because the RPN module can generate better proposals?...

Memory leaks during training!

2

>>> cross_entropy: 0.001210 >>> lr: 0.001000 speed: 2.425s / iter /gruntdata/disk2/hm/CLWSOD/tools/../lib/nets/network.py:569: UserWarning: invalid index of a 0-dim tensor. This will be an error in PyTorch 0.5. Use tensor.item() to convert...