如今我已剑指天涯

About the calculate_class_weighting.py

2

hi, I use the calculate_class_weighting.py generare the loss weight for my training. and I find some problems. First, the weight is： 0.0819 0.4754 0.1324 1.5224 1.5190 2.4730 8.1865 5.2286 0.1870...

update the roadmap

1

hi, I think that the roadmap should add more papers. There are few papers published in 2017.

unsupport multi gpus training?

1

unsupport multi gpus training? I set the gpu_number as 8 and its iteration times is as the same as the gpu_number is 1.

Hi, handong, Could you add the pedestrian trajectory prediction and vehicle trajectory prediction papers? Thx~

Help Stylegan generate better local details

1

Suppose I wanted StyleGAN to focus more on certain areas, such as teeth or mouth, and I added a local discriminator to the mouth when I trained StyleGAN, but it...

Thanks to the excellent work, StyleSpace. However，it is a new concept and one always explores the S space alone. Hence, I create a QQ group (754288725) for communication of StyleSpace....

How combine DynamicRCNN with Cascade-RCNN?

1

Hello, thanks for your great work! How combine DynamicRCNN with Cascade-RCNN? Have you tried it before? Or do you have some parameter configurations to provide?

About the training time

1

Hi, may I bother you about the training time of your method? i.e. how many hours do you spend with how many GPU cards for training the matting network? Looking...