how can i achieve your paper's results evaluation? about deepsegmentor HOT 2 CLOSED

Hi, you can evaluate the models by:

cd eval
python eval.py --metric_mode prf --model_name deepcrack --output deepcrack.prf

You can find some more details in DeepSegmentor/eval.

Then, you can also find the pretrained models for crack detection, I tested the model that achieved the performances listed in the table.

This project is implemented by PyTorch. And the original implementation in our paper is Caffe. There are some differences between the two implementations.

from deepsegmentor.

@chaochon , I updated the codes for improving the performances of PyTorch implementation. You can find the notice here.

from deepsegmentor.