Open nrupatunga opened 6 years ago
@alexgkendall
How is Segnet trained for lower resolution of 224x224 on SUN-RGBD, while the annotations are of arbitrary size. How did you prepare annotation of size 224x224, as I see that resizing the annotation image would disturb the labels?
Is it done considering nearest pixel values?
@nrupatunga Hi, i have the same question about image arbitrary size. Have you solved this problem?
@alexgkendall
How is Segnet trained for lower resolution of 224x224 on SUN-RGBD, while the annotations are of arbitrary size. How did you prepare annotation of size 224x224, as I see that resizing the annotation image would disturb the labels?
Is it done considering nearest pixel values?