Nice Work!!
But I have a question about modality invariant features and modality specific information. What is the meaning of modality invariant features? The mentioned paper [6, 11, 12, 17, 51, 52] extract modality invariant feature only by leveraging modality fusion?
Ai the other hand, what is the modality specific information? Can you give some examples in multi modality Face Anti Spoofing domain?
Nice Work!! But I have a question about modality invariant features and modality specific information. What is the meaning of modality invariant features? The mentioned paper [6, 11, 12, 17, 51, 52] extract modality invariant feature only by leveraging modality fusion? Ai the other hand, what is the modality specific information? Can you give some examples in multi modality Face Anti Spoofing domain?