AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Is There a Problem With Masking?
The training procedure is to minimize the difference of the output of the representation between the clean version and the corrupted version. So, every time you see a percept, you also can perturb it in some way. I think we're going to come up probably in the future with sort of ways to mask that are kind of random, essentially. It's like boss on-machine training. You know, denosinoto encoder method is basically the same thing.