Zero-Shot Visual Recognition using Semantics-Preserving Adversarial Embedding Networks 阅读笔记
Zero-ShotVisualRecognitionusingSemantics-PreservingAdversarialEmbeddingNetworks收获多个网络如何组合?相同目标collaborative,不同目标adversarialtodo查看groudtruthclassembedding查看SAEConclusionprousingadversarialframeworktoco