I want to build a model that when given two vectors, outputs the probability of one vector being the encoded form of the other. I have 2 strategies for this: (Dataset available)
I can directly feed them in concatenation to a neural network and take the output as the probability.
I can train a conditional GAN with the conditional vector being the encoded vector and using the original vector as the generated one. In this case, the discriminator works as the network that I train in the first approach.
Which approach is better? Am I thinking in the right direction?