Cityscapes: PSPNet50 | 0.7730/0.8431/0.9597. | 0.7838/0.8486/0.9617
What is the difference between these two results “Cityscapes: PSPNet50 | 0.7730/0.8431/0.9597. | 0.7838/0.8486/0.9617.“ Why one is higher?
The higher one is tested with multi-scale setting. https://github.com/hszhao/semseg/blob/master/config/cityscapes/cityscapes_pspnet50.yaml#L54
The higher one is tested with multi-scale setting. https://github.com/hszhao/semseg/blob/master/config/cityscapes/cityscapes_pspnet50.yaml#L54
Hi, thanks for your great work, I have a question though, how did you determine the best performance weights file during training?I mean, during training, you validate your model using center crop in each epoch, but not multi-scale validation, does the best performance weights file selected using center crop validation also performs best using multi-scale validation?If not, how do we find the best multi-scale validation weights file?Other than using multi-scale settings in each epoch, cause it's time consuming.Thanks in advance.