Traceback (most recent call last):
File "D:\Projects\remote_sensing_change_detection\train_and_validate.py", line 5, in
from utils.pipeline import RSPipeline
File "D:\Projects\remote_sensing_change_detection\utils\pipeline.py", line 18, in
from models.nested_unet import NestedUNet
ModuleNotFoundError: No module named 'models.nested_unet'
感谢分享源码,初步看了一下,应该是通过分割结果来确定变化区域,但是这种检测方法鲁棒性较差,问下博主是如何优化这个问题的?顺便问下,标注文件都得手动标注吗?是否有公开数据集可以使用的呢?