About phase 2, the shot*3 problem?

[hzhupku]

I found that in the metadata.py, if phase==2, shots=shots*3, is this a bug? I think this will lead to an unfair performance comparison.

[NHW2017]

I found that in the metadata.py, if phase==2, shots=shots*3, is this a bug? I think this will lead to an unfair performance comparison.

Sorry to interrupt you, I have a lot of problems when I run this program. I think the environment I configured is wrong, so can you tell me the environment configuration that you successfully run?

[yanxp]

Hi, in the phase, shots=shots*3 is meaning for base classes to keep the performance of base classes. About the environment, you can refer to the code:https://github.com/jwyang/faster-rcnn.pytorch. Thanks ---------- 该邮件从移动设备发送

…

--------------原始邮件-------------- 发件人："NHW2017 "<[email protected]&gt;; 发送时间：2020年10月19日(星期一) 晚上7:40 收件人："yanxp/MetaR-CNN" <[email protected]&gt;; 抄送："Subscribed "<[email protected]&gt;; 主题：Re: [yanxp/MetaR-CNN] About phase 2, the shot*3 problem? (#51) ----------------------------------- I found that in the metadata.py, if phase==2, shots=shots*3, is this a bug? I think this will lead to an unfair performance comparison. Sorry to interrupt you, I have a lot of problems when I run this program. I think the environment I configured is wrong, so can you tell me the environment configuration that you successfully run? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.

[NHW2017]

thank you for your reply! In fact, I have tried many environment configurations according to the "ReadMe" of the two programs "faster-rcnn.pytorc" and "MetaRCNN", but there is always an error. I don't know where the configuration is correct. If you know it is Please give me some help
Error message "ImportError: /home/xxx/PycharmProjects/MetaR-CNN-master/lib/model/roi_crop/_ext/roi_crop/_roi_crop.so: undefined symbol: __cudaPopCallConfiguration", I also saw the same message in "Issues" , But no one has resolved