Zhengyang Feng

Results 207 comments of Zhengyang Feng

Debug改成True才能用

> ![3](https://github.com/voldemortX/pytorch-auto-drive/assets/114333650/bc2db6c1-fda6-41d5-af36-9a848d3510d1) > 作者您好,请问我用自己拍摄的校内图片resize后做test,.txt中为什么没有任何点的输出呢?请教一下是哪里出了问题呢?命令python main_landet.py --test --config=/home/thesun/PointPainting-master/lineseg/configs/lane_detection/scnn/vgg16_culane.py --checkpoint=/home/thesun/PointPainting-master/lineseg/pt/vgg16_scnn_culane_20210309.pt @thesun9999 这个场景看起来和数据集差异略大,你可以先试下原数据集里的图片拷贝出来能否正常推理,确定是不是操作问题。

> Traceback (most recent call last): > File "main_landet.py", line 75, in > runner = Runner(cfg=cfg) > File "/root/pytorch-auto-drive-3.2.3/utils/runners/lane_det_trainer.py", line 17, in __init__ > super().__init__(cfg) > File "/root/pytorch-auto-drive-3.2.3/utils/runners/base.py", line 120,...

> python main_landet.py --train --config="/root/pytorch-auto-drive-3.2.3/configs/lane_detection/scnn/vgg16_culane.py" --mixed-precision @voldemortX 感觉很奇怪。你的torch是正常能用的吗?

> The difference or significance between Feature Flip Fusion and the middle flip in data augmentation ? Firstly, Feature flip fusion is not a very significant module. It seems to...

> How to use the output of the resnet18s_lstr_tusimple_20210701.onnx? > The input is straightforward i think, but the output is composed of "curves" and "logits" and the got me trying...

@yigitbalcioglu Hi, exactly which config and command did you run?