What's the purpose of the multipication of n_classes to label in evaluation code?

[lx-r]

STDC-Seg/evaluation.py

Lines 56 to 59 in 59ff37f

	hist += torch.bincount(
	label[keep] * n_classes + preds[keep],
	minlength=n_classes ** 2
	).view(n_classes, n_classes).float()

[gongliuqing123]

Hi, are you solve this problem?