Is there any work into building self-verification into these type of networks? For example based on hierarchical categories of concepts?
If part of the network is trained on the concept of a cat, and whether or not an image is a cat is fed into training of the leopard, it seems like the problems would be avoided. Or is the notion that with enough training data and deep enough networks the concept of "leopard is cat" will be learned?
If part of the network is trained on the concept of a cat, and whether or not an image is a cat is fed into training of the leopard, it seems like the problems would be avoided. Or is the notion that with enough training data and deep enough networks the concept of "leopard is cat" will be learned?