Closed chartgod closed 6 months ago
When evaluating performance such as TP, TF, etc., do you evaluate performance by pixel difference between predict_label and correct label?
Is there a source code that evaluates performance among the source codes?
When evaluating performance such as TP, TF, etc., do you evaluate performance by pixel difference between predict_label and correct label?
Is there a source code that evaluates performance among the source codes?