GitHub-Crow

you can see `xx1 = np.maximum(x1[i], x1[order[1:]])` , `x1[order[1:]]` is start from 0, so all the indices in xx1 wiil decrease 1. and `order` is complete (include `i`).

I think there is no obvious difference between them in practice. But in detectron2, there is no +1 operation.