Context Encoders: Feature Learning by Inpainting

1. Encoder-decoder pipeline   Context encoder trained with joint reconstruction and adversarial loss for semantic inpainting Context encoder trained with reconstruction loss for feature learning by fi

>>阅读原文<<