ShirleyHe2020

> 我又用了自己的训练集测试可以了，不用yymnist。 >我用自己的训练集发现报错如下： Traceback (most recent call last): File "make_data.py", line 111, in data[0] = make_image(data, image_paths[idx], ratio) IndexError: list index out of range

I encounter same issue here, can anyone help fix this ?

> when i run python tools/train_net.py, there is a error: > from fcos_core.utils.env import setup_environment > ModuleNotFoundError: No module named 'fcos_core' Did you get this solved ?

> Did you install the project? Please follow the instructions here https://github.com/tianzhi0549/FCOS/blob/master/INSTALL.md#option-1-step-by-step-installation cuda 10.1 in my environment, so I run ' conda install pytorch torchvision cudatoolkit=10.2 -c pytorch' step by...

> I'm wondering whether BBN coule be introduced into one-stage object detection method, such as retinanet and yolo. Any improvement could be obtained ? Have you guys try? > >...

Hi Authors, can this work be combined into YOLO v5 ？

same issue here every time I run make_data.py, the images were removed in dir Image