AI Paper Atlas

Research map / Label-efficient and robust learning / Noisy labels

HaNa: Hardness and Noise-Aware Robust Cross-modal Retrieval

Fangming Zhong, Haiquan Yu, Cun Zhu, Suhua Zhang

AAAI 2026 · Label-efficient and robust learning › Noisy labels · approach: Noisy cross-modal hashing and multi-label · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add weighted contrastive loss (vs standard contrastive loss) (ablation)HaNa · Flickr30K 40% noiseimage-to-text R@1: 76.0 -> 88.5Improved
add weighted contrastive loss (ablation)HaNa · Flickr30K 40% noisetext-to-image R@1: 57.5 -> 75.1Improved
add momentum update of weights (ablation)HaNa · Flickr30K 40% noiseimage-to-text R@1: 87.6 -> 88.5Improved
add momentum update of weights (ablation)HaNa · Flickr30K 40% noisetext-to-image R@1: 74.1 -> 75.1Improved
add hardness weights on clean samples (vs weight 1) (ablation)HaNa · Flickr30K 40% noiseimage-to-text R@1: 86.5 -> 88.5Improved
add hardness weights on clean samples (ablation)HaNa · Flickr30K 40% noisetext-to-image R@1: 73.5 -> 75.1Improved
add ANRL (ablation)HaNa · Flickr30K 40% noiseimage-to-text R@1: 87.3 -> 88.5Improved
add ANRL (ablation)HaNa · Flickr30K 40% noisetext-to-image R@1: 74.3 -> 75.1Improved
HaNa versus NPC at 60% noiseHaNa · Flickr30K 60% noiseimage-to-text R@1: 83.0 -> 87.1Improved
HaNa versus NPC on real-world noiseHaNa · CC120KRSUM: 517.6 -> 521.5Improved
increase noise ratio from 20% to 60% for HaNaHaNa · Flickr30Kimage-to-text R@1: 89.0 -> 87.1Mixed
HaNa versus NPC on MS-COCO text-to-image R@1 at 20% noiseHaNa · MS-COCO 1K, 20% noisetext-to-image R@1: 66.3 -> 69.0Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy