tigerzjh

In the prototxt, the group number of the 3*3 conv layer is not set. However i am troubled whether the number of group should be 3 or the number of...

In the prototxt, the group number of the 3*3 conv layer is not set. However i am troubled whether the number of group should be 3 or the number of...

Tag Augmentation of Images

5

Thank you for the great work! 1、The paper says Note that each image from ImageNet-11K is annotated by a single tag. Is each image just annotated by the leaf tag?...

is se better for the performance？

1

hi，the green line which is means use se-yolov5+nst is likely better than the pink line which is just use se-yolov5+nst?

can you share the code of use grad cam on our model in the paper?

2

i am troubled how to use grad cam in se-nets, for there is not a conv layer to show like others.

Mac press button， the button color change slowly even nothing。

1

in Mac，when i click the label button， the color of the label button does not change。

[BUG] import nvcv cvcuda 失败

1

**Describe the bug** 使用tar 方式安装，由于权限原因无法使用deb方式。对应的路径如下：  我是将这个路径下的so，放入环境变量吗？ export PYTHONPATH=/home/zhangjianhu/softwars/opt/nvidia/cvcuda0/lib/x86_64-linux-gnu:$PYTHONPATH 之后仍然import失败