AI Paper Atlas

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

Multi-modal Identity Extraction

Ryan Webster, Teddy Furon

ICCV 2025 · Vision-language and multimodal › Cross-modal learning · approach: Multi-modal object re-identification · 0 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
use full Wikipedia name set as distractors (S3) instead of S1IMIA · ViT-G-14 L2BTPR@FPR=0.25%: 16.7 -> 34.6Improved
use full Wikipedia name set (S3) instead of S1IMIA · ViT-H-14 L2BTPR@FPR=0.25%: 19.6 -> 36.4Improved
use full Wikipedia name set (S3) instead of S2IMIA · ViT-B-32 L2BTPR@FPR=1%: 57.2 -> 52.8Hurt
add adaptive gap between top-2 frequenciesIESA · Convnext-xxlarge L2BE@99.5: 0.70 -> 15.85Improved
add adaptive gap between top-2 frequenciesIESA · Evae-14 L2BE@99.5: 0.68 -> 21.78Improved
add adaptive gap between top-2 frequenciesIESA · ViT-G-14 L2BE@95: 52.04 -> 47.42Hurt
limit attacker to 10 images per identity instead of all (ablation)IESA adaptive · ViT-G-14 L2BE@99.5: 32.53 -> 18.65Hurt
limit attacker to 10 images per identity (ablation)IESA adaptive · Evae-14 L2BE@99.5: 21.78 -> 0.03Hurt
append 10k distractor names (IESA step) to generated names (ablation)IEGA · VLM generationsprecisionImproved
use GPT-4o instead of REINFORCE for name generationIEGA · VGGFace2 4000 identitiestotal extraction rate: 12.68 -> 52.51Improved
use Pixtral-12B instead of REINFORCE (ViT-L-14 L2B)IEGA · VGGFace2 4000 identitiestotal extraction rate: 12.68 -> 8.40Hurt
apply strict face crop to attacker images (ablation)IESA · ViT-G-14 L2BE@95: 44.78 -> 25.91Hurt

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