yudawen

pytorch版本

2

你好！我想知道你对于作者原版的代码做了哪些修改，使得它能够在pytorch1.3和torchvision0.4.1上工作呢？我看到作者现在的代码要求torch>1.5。

Object detection

1

Hello! Have you implemented object detection using spherical convolution? I have some questions to ask you.

input size

1

I want to know the aspect ratio of an equirectangular image should be 1:2, why the author set the Omni-MNIST with a size of 60×60. For spherical convolution, can the...

using resnet 50

1

Hello! When I use resnet50 instead of vgg16, the loss of the second stage box and the loss of the category are always small, and the final result is not...

What number of targets corresponding to the original image can be detected?

Does it support PyTorch1.0 or higher?

1

Hello, this branch supports PyTorch1.0 or higher now?