Xu Hai

VESPCN is hard to understand, especially the motion compensation model.

difference between macro USE_METAL, CAFFE2_USE_MPSCNN & CAFFE2_MOBILE

7

For iOS system, do I need to enable all these macros? USE_METAL, CAFFE2_USE_MPSCNN CAFFE2_MOBILE

when will the joint fusion model (Coming soon) come?

1

I want to evaluate the joint model. Thanks.

I have converted to python, but can only input 1025*2049 size image

1

I write a python script to test the pre-trained model. But, I can only do test on resolution of 1025*2049 images, other resolution does not work.

How to training?

41

Using this: PSPNet/evaluation/prototxt/pspnet101_VOC2012_473.prototxt change the data layer to the training data?

Take pspnet101_VOC2012_473.prototxt as an example, the input image resolution should be 473*473. When I try to call: `net.blobs['data'].reshape(1, *in_.shape) ` This will result error

what is Levin09.mat? how to create Levin09.mat?

running with FCN.py, but the result not so good

3

This is my predicted matting. Any suggestions? 

I am compress the https://github.com/BVLC/caffe/tree/master/models/bvlc_alexnet and decompress it, then use this decompressed model file for testing. I have tested several images, everything works well except this image: caffe/examples/images/cat.jpg (https://github.com/BVLC/caffe/blob/master/examples/images/cat.jpg) give...

what's wrong? And training exit with only 20 steps??