AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

Cross-modal Fuzzy Alignment Network for Text-Aerial Person Retrieval and A Large-scale Benchmark

Yifei Deng, Chenglong Li, Yuyang Zhang, Guyue Hu, Jin Tang

CVPR 2026 · Vision-language and multimodal › Cross-modal learning · approach: Probabilistic embeddings for cross-modal retrieval · 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 Context-Aware Dynamic Alignment (CDA) loss (ablation)bridge-alignment baseline · AERI-PEDESRSum: 174.84 -> 183.04Improved
add CDA loss (ablation)bridge-alignment baseline · AERI-PEDESRank-1: 43.88 -> 46.18Improved
add Fuzzy Token Alignment (FTA) (ablation)bridge-alignment baseline · AERI-PEDESRSum: 174.84 -> 176.64Improved
add FTA on top of CDA (ablation)CDA model · AERI-PEDESRSum: 183.04 -> 186.65Improved
add FTA on top of CDA (ablation)CDA model · AERI-PEDESmAP: 43.98 -> 44.79Improved
use aerial images as the CDA bridge instead of none (ablation)CFAN · AERI-PEDESRSum: 182.80 -> 184.01Improved
use ground images as the CDA bridge instead of none (ablation)CFAN · AERI-PEDESRSum: 182.80 -> 186.65Improved
use ground images rather than aerial images as bridge (ablation)CFAN · AERI-PEDESRSum: 184.01 -> 186.65Improved
add ground-view images as auxiliary cuesCFAN · TBAPRRSum: 188.50 -> 189.03Improved
add ground-view images as auxiliary cuesCFAN · TBAPRRank-10: 73.47 -> 73.06Hurt
increase sensitivity k in CDA from 1 to 16 (ablation)CDA module · AERI-PEDESRSum: 186.65 -> 183.31Hurt
compare CFAN with ground images against HAMCFAN vs HAM · AERI-PEDESRSum: 180.77 -> 186.65Improved

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