AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Multi-Spectral Vehicle Re-Identification: A Challenge

Hongchao Li, Chenglong Li, Xianpeng Zhu, Aihua Zheng, Bin Luo

AAAI 2020 · Label-efficient and robust learning › Recognition and long-tail · approach: Supervised person re-identification · 90 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 heterogeneous score coherence loss to MSN (ablation)MSN (ResNet50) · RGBN300mAP: 56.9 -> 59.1Improved
add class-aware ID loss to MSN (ablation)MSN (ResNet50) · RGBN300mAP: 56.9 -> 59.4Improved
add both losses to MSN (full HAMNet) (ablation)MSN (ResNet50) · RGBN300mAP: 56.9 -> 61.9Improved
add both losses to MSN (full HAMNet) (ablation)MSN (ResNet50) · RGBNT100mAP: 60.5 -> 64.1Improved
add heterogeneous score coherence loss to MSN (ablation)MSN (ResNet50) · RGBNT100mAP: 60.5 -> 61.4Improved
add class-aware ID loss to MSN (ablation)MSN (ResNet50) · RGBNT100Rank-1: 82.6 -> 83.1Improved
add NIR stream to RGBResNet50 IDE · RGBN300mAP: 49.5 -> 56.9Improved
add TIR stream to RGB-NIRMSN (ResNet50) · RGBNT100mAP: 43.1 -> 60.5Improved
replace MSN with HAMNetDensenet121 · RGBN300mAP: 46.1 -> 46.7Improved
use HAMNet instead of extended person Re-ID methods (PCB, MGN, ABD)ResNet50 · RGBNT100Rank-1: 85.1 -> 84.7Mixed

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