seen-unseen prior

semantic descriptor (1) logit adjustment (1) generalized zero-shot learning (1) semantic embedding (1) generative model (1) variational bayesian inference (1) variational inference (1) zero-shot learning (1)