zeyang

Report a Bug

1

Hi, I think there might be a bug here, a strange break makes it only traverse one image before it get out of the for loop. Could you please have...

This aims to fix issue #17. The compling error is actually caused by missing the file nms_cuda.c. I borrow the one in [Detectron.pytorch](https://github.com/roytseng-tw/Detectron.pytorch/blob/master/lib/model/nms/src/nms_cuda.c). close #17

Compiling Error

1

When I compile the lib with cuda-9.0, it poped up an error shown in [log.txt](https://github.com/fanq15/FSOD-code/files/4730450/log.txt). Could you advise? Thanks.

Full Results on ADE20k

3

Hi, may I have the class-wise results for ADE20k? Thanks.

Reproduced results lower than the reported ones

2

Hi, I directly ran the released codes without any modification. However, I found that the obtained results are lower than the reported ones by >1 percent point, especially the 10-1...

Could you provide the base mAP of COCO 10-shot and 30-shot? or anybody who has run this code could share the results. Many thanks.