I'm sorry to disturb you

[zhoujh2020]

I want to reproduce the FCN model in your paper, but the training has been unsuccessful.Can I ask how you train the model, such as the pre training model and the changes to the protoxt file.I would appreciate it if you could remember and tell me the details.Tank you very much.

[fyangneil]

@zhoujh2020 If I remember correctly, we use Caffe version FCN https://github.com/shelhamer/fcn.berkeleyvision.org, you can try to initialize the model with the pretrained model on ImageNet, and try different setting like FCN-8s, 16s, or 32s to see which one can work. In my mind, I think I use the 8s, all the other settings are simply as default.