danfenghong
commented
2 years ago Hi, The shape of the feature you get is only for the testing samples, i.e., Houston 2013 dataset, you can refer to our paper, you can find the testing samples have 12197. Note that this is not a classification map. It is just the classification results for the 12197 testing samples to compute the classification accuracy, e.g., OA, AA, and kappa. If you would like to generate the classification map, you will feed all samples in the whole image into the trained model and get the prediction of all samples. Perhaps, you can refer to our another code: https://github.com/danfenghong/IEEE_TGRS_CCR-Net to see how to get the whole classification map.
Cheers, Danfeng
trilobiteinCambrian @.***> 于2022年4月12日周二 21:18写道：
Hi Danfeng can you introduce how to get the classification results of ENDNet？I just got the result（“feature.mat”）. and the shape of feature is 15*12197，（I cannot understand what the feature means）how can i transform it to the shape of original image and feed into SVM？Or it cannot reshape to the shape of orginal image？as a beginner，I have a lot of questions，please forgive me，thanks.
— Reply to this email directly, view it on GitHub https://github.com/danfenghong/IEEE_GRSL_EndNet/issues/1, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFL2GZR6US6DCGZPVPPLUQLVEVZZZANCNFSM5THJIGPQ . You are receiving this because you are subscribed to this thread.Message ID: @.***>
trilobiteinCambrian
Shahrzad92
W2087838645
Hi Danfeng can you introduce how to get the classification results of ENDNet？I just got the result（“feature.mat”）. and the shape of feature is 15*12197，（I cannot understand what the feature means）how can i transform it to the shape of original image and feed into SVM？Or it cannot reshape to the shape of orginal image？as a beginner，I have a lot of questions，please forgive me，thanks.