Intra-Modal Neighbors Never Lie: Rectifying Inter-Modal Noisy Correspondence via Graph-Based Intra-Modal Reasoning
About
Large-scale web-harvested datasets have fueled the progress of cross-modal retrieval but inevitably suffer from noisy correspondence, which severely degrades model generalization. Existing methods primarily address this by filtering out noise or seeking a substitute label, yet they predominantly remain bound by a "Discrete Selection" paradigm. We argue that relying on a single discrete proxy induces Single-Point Fragility and Discretization Error. To overcome these limitations, we propose a novel framework, Intra-modal Neighbor-aware Noise Rectification (IN2R), which shifts the paradigm from searching for a substitute to synthesizing a reliable supervision target. Leveraging the intrinsic geometric stability of intra-modal data, IN2R employs a Graph Refiner to perform relational reasoning over neighbors retrieved from a dynamic Cross-Model Memory. Instead of propagating discrete labels, our method synthesizes a continuous, soft prototype that reflects the consensus of the local semantic neighborhood, effectively rectifying inter-modal misalignment. Extensive experiments on Flickr30K, MS-COCO, and CC152K demonstrate that IN2R significantly outperforms state-of-the-art methods. Our code and pre-trained models are publicly available at https://github.com/liuyyy111/IN2R.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Text Retrieval | Flickr30K 1K (test) | R@180 | 115 | |
| Image Retrieval | Flickr30K 1K (test) | R@160.5 | 103 | |
| Image Retrieval | MSCOCO 5-fold 1K (test) | Recall@164.2 | 51 | |
| Text Retrieval | MS-COCO 5-fold 1K (test) | R@181.6 | 33 | |
| Image-Text Retrieval | CC152K (test) | Rsum380.8 | 19 | |
| Text Retrieval | CC152K 150K (test) | R@145.2 | 9 | |
| Image Retrieval | CC152K 150K (test) | R@143.3 | 9 |