Semi-Supervised Cross-Modal Retrieval Based on Discriminative Comapping

Figure 6

Image queries and the corresponding top retrieved images, and text queries and the corresponding top retrieved images.