tztztztztz / yolov2.pytorch Goto Github PK

Hi,
I know that this repo works with a squared image. But, I am trying to incorporate a rectangle image of size (720,1440).
However, I end up with an error in the below line as it is not able to match the expected size.

could you please tell me, how did you come up with these params? If I have to use a rectangle image what should be changed and on what basis?
Thanks in advance

Hi. Did you reproduce the results of this paper which were the same as the original paper? I'm the beginner too, I want to try to reproduce it. Thanks.

thank you for your good code
training was ok
test was ok
how to get mAP?

this step, there are some errors, I searched it but there are no effective solution, I tried some ways by changing the path, still didn`t work?
ln -s $VOCdevit/VOC2007 VOC2007
error:
ln: failed to creat symbolic link 'VOC2017': operation not supported.

Hi, thank you for such great work, but I want to confirm that it is a Yolov2 network like the one in the original paper? Because when I look at your code, I think it seems like the Yolov1:), but you know I'm also a beginner in deep learning, so I just want to confirm this with you. Thank you!

Hello, I want to triain my own dataset instead pascal VOC dataset.
How can I change your code?

hello, tztztztztz:
thankyou for your opensource code, it's very good!
but i have a question about loss, i read some blog about the loss, the yolov2 loss is like the picture, two red box in the picture is not find in your loss.py, is i didn't find or you do not write?

thank you for your good code would you please guide me for how to calculate iou?
again thank you

Thanks a lot for the great repo.
I am trying to use the repo with custom dataset. I have my own input pipeline which provides the required data. However, I have few images where there are no bboxes. I would like to know how to handle this situation? Do you handle it somewhere in the repo?