參考文章：https://gitmemory.com/issue/aim-uofa/AdelaiDet/235/731017126
出現原因
detectron2會不斷有一些更新，這就會導致BlendMask作者用的detectron2和我們自己裝的detectron2不完全一致，這就是報錯的原因，
解決方法
將BlendMask源代碼中train_net.py檔案第128行的build_detection_train_loader(cfg, mapper)修改為build_detection_train_loader(cfg, mapper=mapper)，修改之后這一段代碼為：
def build_train_loader(cls, cfg):
"""
Returns:
iterable
It calls :func:`detectron2.data.build_detection_train_loader` with a customized
DatasetMapper, which adds categorical labels as a semantic mask.
"""
mapper = DatasetMapperWithBasis(cfg, True)
return build_detection_train_loader(cfg, mapper=mapper)
根據我的經驗，處理完這個問題之后后續還會有類似的報錯，如__init__() takes 1 positional argument but 4 were given等，暫時倒還沒有什么好的想法可以處理，
本來我已經放棄了跑BlendMask的，結果在評論區被追問出現__init__() takes 1 positional argument but 4 were given怎么辦，只好硬著頭皮研究了一下，，，
解決方法
將train_net.py第69行的super(DefaultTrainer, self).__init__(model, data_loader, optimizer)換成super(Trainer, self).__init__(cfg)即可，
轉載請註明出處，本文鏈接：https://www.uj5u.com/houduan/236080.html
標籤：python