Closed-form Sample Probing for Learning Generative Models in Zero-shot Learning

The proposed sample probing approach improves the ZSL results even when integrated into state-of-the-art generative models, and opens doors to end-to-end training thanks to the differentiability of the solvers.