PSPNet

PSPNet copied to clipboard

I have a question, why design four layers with 1*1, 2*2, 3*3, 6*6 pool layer, instead of 1*1, 2*2, 3*3, 8*8

cvlearn913

92ypli

How to initialize model when training with other class

2

Hi @hszhao and all. I want to train with other class. Ex: Clothes, hat, ball....... So where can i find file .cafemodel(Pre-trained models) for initialize?

ThienAnh

eval_acc.m

2

qilinli

Questions about PSPNet.

6

Thank you for uploading your code. It is very helpful to understand PSPNet. I have two questions about your paper. 1. You wrote > we use a pretrained ResNet model...

kazucmpt

Training accuracy is stuck at 0

27

@hszhao I am training PSPNet with my own dataset of image size 360X480 and total three class. I have limited GPU memory (12 GB) and had to keep the batch...

bhadresh74

Reason to have a fixed inference size (473x473)

1

Hi, Thank you for sharing the code and trained models! I have a question specific to the demo in your PyTorch implementation. As I understand, you are using a base_size...

TouqeerAhmad

Can I use the pre-training model to calculate the miou and mean Pixel Accuracy of a small-sized image, such as 256*256

mxkai

Hi, GPU GTX 1070 8G memory is not enough?

1

Hi, thanks for your work, I use GTX 1070 8G memory to run the eval_all.m. But it tell me out of memory. Is it not enough? The error is： processing...

Leequens

From which model did you fine-tune?

3

Hi, you said you use resnet-101 with dilated convolution, but I can't find any pre-trained ResNet with dilated convolution on the Internet. I want to know more in details.

shansiliu