SamsungLabs / fbrs_interactive_segmentation

[CVPR2020] f-BRS: Rethinking Backpropagating Refinement for Interactive Segmentation https://arxiv.org/abs/2001.10331

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Any reason behind to set crop_size to (320, 480)?

MaitaYuki opened this issue · comments

Dear authors,

Thank your for sharing your great work.
I want to train your method on my own dataset. The image resolution in my dataset is 1280 * 760. In the line 29 of r34_dh128.py, model_cfg.crop_size = (320, 480). What is the reason behind to set crop size to (320, 480)? Do I need to set to a bigger window since the image resolution is bigger?
Many thanks!

Hi, we chose that crop size mainly due to memory and speed limitations during training (anyway, it's more or less comparable to the image sizes from training datasets). Of course, you have to adapt the crop size to your dataset and hardware to achieve better performance.