上一条: Shadow Generation for Composite Image in Real-world Scenes
下一条: Adversarial Query-by-Image Video Retrieval Based on Attention Mechanism