AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

CodeMamba: Shifting from Target Semantics to Self-Supervised Background Manifold Learning for Singularity Detection in Infrared Sequences

Jingwen Ma, Xinpeng Zhang, Fan Shi, Xu Cheng, Shengyong Chen

ICML 2026 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Pretext-task visual pre-training

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 Bayesian uncertainty-weighted fusion (ablation)CodeMamba (GBAPM+MHSFM) · IRDSTmAP50: 77.35 -> 78.24Improved
add Bayesian uncertainty-weighted fusion (ablation)CodeMamba (GBAPM+MHSFM) · IRDSTF1: 88.28 -> 90.62Improved
add Bayesian uncertainty-weighted fusion (ablation)CodeMamba (GBAPM+MHSFM) · DAUBmAP50: 96.74 -> 97.42Improved
add Bayesian uncertainty-weighted fusion (ablation)CodeMamba (GBAPM+MHSFM) · DAUBrecall: 98.74 -> 99.75Improved
add MHSFM to GBAPM (ablation)CodeMamba · IRDSTmAP50: 76.31 -> 77.35Improved
add GBAPM to MHSFM (ablation)CodeMamba · IRDSTmAP50: 76.86 -> 77.35Improved
add MHSFM to GBAPM (ablation)CodeMamba · DAUBmAP50: 96.32 -> 96.74Improved
add GBAPM to MHSFM (ablation)CodeMamba · DAUBmAP50: 96.05 -> 96.74Improved
add GBAPM to MHSFM (ablation)CodeMamba · DAUBF1: 98.36 -> 98.36No clear effect
add frequency path to spatial path in HSF (ablation)MHSFM · IRDSTmAP50: 75.80 -> 78.24Improved
add spatial path to frequency path in HSF (ablation)MHSFM · IRDSTmAP50: 73.33 -> 78.24Improved
use bidirectional instead of forward-only scanning (ablation)HSF · IRDSTmAP50: 72.15 -> 78.24Improved

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