Hello, I really like this dataset and I would like to convert my own data to the format of the CitySpaces dataset. I only have two classes, and I have modified the labels as follows:
labels = [ # name id trainId category catId hasInstances ignoreInEval color Label( 'unlabeled' , 0 , 255 , 'void' , 0 , False , True , ( 0, 0, 0) ), Label( 'road' , 1 , 0 , 'flat' , 1 , False , False , (128, 64,128) ), Label( 'pudele' , 2 , 1 , 'flat' , 1 , False , False , (244, 35,232) ), ]
However, when I train with improved data, I always make mistakes, such as failing to improve metrics or predicting tensors that are all 0 when inferring images, which means all predictions are for the road category.
May I ask if I need to change any other parameters when converting data? Or was it an error in training the model's network structure or reading data?
Looking forward to your reply. Thank you in advance.
Hello, I really like this dataset and I would like to convert my own data to the format of the CitySpaces dataset. I only have two classes, and I have modified the labels as follows:
labels = [ # name id trainId category catId hasInstances ignoreInEval color Label( 'unlabeled' , 0 , 255 , 'void' , 0 , False , True , ( 0, 0, 0) ), Label( 'road' , 1 , 0 , 'flat' , 1 , False , False , (128, 64,128) ), Label( 'pudele' , 2 , 1 , 'flat' , 1 , False , False , (244, 35,232) ), ]However, when I train with improved data, I always make mistakes, such as failing to improve metrics or predicting tensors that are all 0 when inferring images, which means all predictions are for the road category.
May I ask if I need to change any other parameters when converting data? Or was it an error in training the model's network structure or reading data?
Looking forward to your reply. Thank you in advance.