Fig. 6
From: Dynamic adaptive synergistic attention network for visible-infrared person re-identification
Top-10 retrieval examples from multiple datasets and modalities. For each query (left), the ten most similar gallery images are ranked left to right. Green borders indicate correct matches; red borders indicate incorrect matches. These results illustrate the model’s ability to match identities across modalities, even under challenging conditions like low illumination in the LLCM dataset.
