AI Paper Atlas

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

Text-based Aerial-Ground Person Retrieval

Xinyu Zhou, Yu Wu, Jiayao Ma, Wenhao Wang, Min Cao, Mang Ye

AAAI 2026 · Vision-language and multimodal › Cross-modal learning · approach: Probabilistic embeddings for cross-modal retrieval · 6 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 HR-MoE blocks to image encoder (ablation)TBPS-CLIP baseline · TAG-PEDESR@1: 59.09 -> 61.60Improved
use vanilla MoE instead of ViT block (ablation)TBPS-CLIP baseline · TAG-PEDESR@1: 59.09 -> 60.26Improved
hierarchical routing and expert grouping vs vanilla MoE (ablation)TAG-CLIP · TAG-PEDESR@1: 60.26 -> 61.60Improved
add orthogonal loss on top of HR-MoE + view loss (ablation)TAG-CLIP · TAG-PEDESR@1: 61.55 -> 62.64Improved
add view loss on top of HR-MoE + orthogonal loss (ablation)TAG-CLIP · TAG-PEDESR@1: 62.32 -> 62.64No clear effect
add view and orthogonal losses to HR-MoE (ablation)TAG-CLIP · TAG-PEDESmAP: 48.94 -> 49.27No clear effect
train on aerial+ground instead of ground only (ablation)TAG-CLIP · TAG-PEDES subset, ground testR@1: 63.57 -> 64.49Improved
train on aerial+ground instead of aerial only (ablation)TAG-CLIP · TAG-PEDES subset, aerial testR@1: 54.04 -> 54.64Improved
train on aerial+ground instead of aerial only (ablation)TAG-CLIP · TAG-PEDES subset, aerial testR@5: 73.74 -> 73.01Hurt
TAG-CLIP vs RDE on TAG-PEDESRDE · TAG-PEDESR@1: 61.58 -> 62.64Improved
TAG-CLIP vs RDE on ground-only CUHK-PEDESRDE · CUHK-PEDESR@1: 75.94 -> 74.38Hurt
TAG-CLIP vs TBPS-CLIP on ground-only CUHK-PEDESTBPS-CLIP · CUHK-PEDESR@1: 72.66 -> 74.38Improved

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