Zero-shot learning on the ImageNet of over 20000 classes

Devise: A deep visual-semantic embedding model (NIPS2013)

Zero-Shot Learning by Convex Combination of Semantic Embeddings (http://arxiv.org/pdf/1312.5650.pdf)

Mohammad Norouzi, Tomas Mikolov, Samy Bengio, Yoram Singer, Jonathon Shlens, Andrea Frome, Greg S. Corrado, Jeffrey Dean

Unsupervised learning of Neural Network outputs

Yao Lu

1) A visual representation of classes (Inspire by Hinton's Dark knowledge)

2) a semantic representation of classes from the WordNet graph,

3) a bridge between the visual and semantic representations.