Keep unique pairs of values from two vectors in numpy

Suppose I have two integer numpy vectors a0 and a1 of the same length. I want to return unique pairs of elements of a0 and a1. For instance:

a0 = np.asarray([0, 1, 4, 1, 0, 4]) # --> b0 = np.asarray([0, 1, 4, 4]) a1 = np.asarray([0, 2, 3, 2, 0, 2]) # --> b1 = np.asarray([0, 2, 3, 2])

What is the most concise way to do this in numpy?