Learning to Rematch Mismatched Pairs for Robust Cross-Modal Retrieval

Benchmark Model Rank Results
cross-modal-retrieval-with-noisy-1L2RM-SGRAF#3R-Sum: 374.2Image-to-text R@1: 43.0Image-to-text R@5: 67.5
cross-modal-retrieval-with-noisy-2L2RM-SGRAF#6R-Sum: 503.8Image-to-text R@1: 77.9Image-to-text R@5: 95.2
cross-modal-retrieval-with-noisy-3L2RM-SCARF#6R-Sum: 524.7Image-to-text R@1: 80.2Image-to-text R@5: 96.3