Jaemin Jung

@yang8595 @huidemie I'm sorry keep your waiting. test.py is renamed [inference.py](https://github.com/jaemin93/pytorch-deeplab-xception/blob/master/inference.py)

@brx290 0. if you can find experiment_{0, 1, ...} folder, there is checkpoint file (*.pth.tar). 1. Then, you define deeplabv3 model (backbone xception) and load checkpoint file. 2. Just give...

@brx290 No, you don't have to. you make inference.py ``` # define model = DeepLab(num_classes=self.nclass, backbone=args.backbone, output_stride=args.out_stride, sync_bn=False, freeze_bn=False) # load checkpoint = torch.load('your/pth.tar/file', map_location='cpu') model.load_state_dict(checkpoint['state_dict']) # prepare inference model.eval()...

@brx290 you need to map the predicted values ​​to RGB values. [[0, 0, 0], [0, 1, 0], [0, 1, 1]] There is no difference between 0 and 1 in human...

@brx290 That's because the file(summaries.py) imports the tensorboardX package. It would be helpful to refer to this. https://pytorch.org/docs/stable/tensorboard.html It's simple and you'll understand right away. Then I hope you enjoy...

Test.py is renamed inference.py 보낸 사람: zhujiesuper 회신 대상: jfzhang95/pytorch-deeplab-xception 날짜: 2020년 4월 29일 수요일 오후 2:22 받는 사람: jfzhang95/pytorch-deeplab-xception 참조: Jaemin jung , Mention 주제: Re: [jfzhang95/pytorch-deeplab-xception] test.py (#122)...

I have same problem. I saved ckpt and load ckpt to inference. but its result is very different [train->prediction]. i check test data and find what is problem. INPUT_SIGNAL_TYPES is...

https://github.com/mAy-I/deepstream_centernet I made it here !