Mr. Long
Mr. Long
嗯嗯，在我这里跑的时候前面都会有一个“backbone.model.”，这个在下载的模型里是没有的。 不过，项目里efficientne-tb2-voc.pth这个权重是可以直接加载的
在 train.py 里 `model = YoloBody(Config, phi=phi, load_weights=False)` 把load_weights 改成True 然后注释掉下面的权重加载那一块就可以了
https://github.com/Alioth5670/fog_simulation.git 我之前写过一个，应该能用吧
hello @gdaddi. I followed the same steps as you to complete the code migration in 5.3, but I could only capture the color image, and there was only one empty...
I solved this problem by saving images in RGBA format, which had some problems displaying images on my computer, and using python to read them was no problem