Regarding training the model on custom dataset for ariel images on binary segmentation.

[goku206125]

问题确认 Search before asking

• [X] 我已经搜索过问题，但是没有找到解答。I have searched the question and found no related answer.

请提出你的问题 Please ask your question

Can you please tell me how I should make my own dataset config file to use paddleseg to train model on my own dataset.

I have to segment two classes ( "road' and "not road") from ariel images. I have following the custom dataset instructions but its not helping.

I have used the custom dataset tutorial ( python tools/data/split_dataset_list.py D:\PaddleSeg\data D:\PaddleSeg\data\images D:\PaddleSeg\data\labels --split 0.6 0.2 0.2 --format "jpg" "jpg") to split my dataset.

But after changing the parameters of model, it is getting error File "C:\Users\nn\miniconda3\envs\paddleseg\lib\site-packages\paddle\nn\functional\loss.py", line 2818, in cross_entropy _, out = _C_ops.cross_entropy_with_softmax( ValueError: (InvalidArgument) Input(Logits) and Input(Label) should in same shape in dimensions except axis. [Hint: Expected logits_dims[i] == labels_dims[i], but received logits_dims[i]:512 != labels_dims[i]:3.] (at ..\paddle\phi\infermeta\binary.cc:940).

config_paddle.txt