SCNN_Pytorch
SCNN_Pytorch copied to clipboard
harryhan618
Pytorch implementation of "Spatial As Deep: Spatial CNN for Traffic Scene Understanding"
when I run "make", I met the following problem: Scanning dependencies of target evaluate [ 20%] Building CXX object CMakeFiles/evaluate.dir/src/evaluate.cpp.o [ 40%] Building CXX object CMakeFiles/evaluate.dir/src/counter.cpp.o [ 60%] Building CXX...
Hello, is "hlight" in the evaluation result the same as "Dazzle light" in the paper？
请问 我用您提供的exp0_best.pth去测试单张图片报错了 size mismatch for fc.0.weight: copying a param with shape torch.Size([128, 2880]) from checkpoint, the shape in current model is torch.Size([128, 4500]).我该怎么办， 请指教
您好，在训练时，CUlane.py,line 48 "img=cv2.cvtcolor(img,cv2.COLOR-BGR2RGB)"出现了cv2.error:opencv(3.4.4)/io/opencv/modules/imgproc/src/color.cpp：182：error：（-215：assertion failed）！-src.empty()in function 'cvtcolor'的错误,网上大多说是路径问题，但是我实在看不出我的路径有什么问题，请问您可以帮我解答一下吗
您好，我有2个问题想请教。 1）请问为何在val模式下， Loss的值没有求平均呢，即：最后在tensorboard上显示的loss值。 ` val_loss += loss.item() val_loss_seg += loss_seg.item() val_loss_exist += loss_exist.item() progressbar.set_description("batch loss: {:.3f}".format(loss.item())) progressbar.update(1) progressbar.close() iter_idx = (epoch + 1) * len(train_loader) # keep align with training...
I modify all configs according to my requirements. But It outputs "Segmentation fault" without any other information. So can you help me solve this? Very thanks.
I follow your advice in the net ,and training CULane data,and I have download your best exp10_best.pth ,and put it in the ./experiments/exp10 folder ,and then I in put the...
train.py can run successfully, but when it comes to test, it raise the 'Some raw_file from your predictions do not exist in the test tasks'. How should I solve this...
Exception: Some raw_file from your predictions do not exist in the test tasks. 有人遇到过这样的问题嘛？
Metadata
Owner
Metadata
Pytorch implementation of "Spatial As Deep: Spatial CNN for Traffic Scene Understanding"