AI Paper Atlas

Research map / Vision-language and multimodal / CLIP and zero-shot

Temporal Inversion for Learning Interval Change in Chest X-Rays

Hanbin Ko, Kyeongmin Jeon, Doowoong Choi, Chang Min Park

CVPR 2026 · Vision-language and multimodal › CLIP and zero-shot · approach: Self-distillation for vision-language pretraining · 2 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 Change-aware Sigmoid pretraining loss on inverted pairs (ablation)BioViL-T SigLIP · MS-CXR-T supervisedStandard avg macro-accuracy: 61.1 -> 64.6Improved
add Change-aware Sigmoid pretraining loss (ablation)BioViL-T SigLIP · MS-CXR-T supervisedConsistency avg: 39.5 -> 45.4Improved
add Bidirectional Cross-Entropy (BiCE) in fine-tuning (ablation)BioViL-T + Change-aware SigLIP · MS-CXR-T supervisedConsistency avg: 45.4 -> 53.1Improved
add BiCE in fine-tuning (ablation)BioViL-T + Change-aware SigLIP · MS-CXR-T supervisedStandard avg: 64.6 -> 65.2No clear effect
add Temporal Consistency Loss (TCL) on top of BiCE (ablation)BioViL-T · MS-CXR-T supervisedConsistency avg: 53.1 -> 57.4Improved
add TCL on top of BiCE (ablation)BioViL-T · MS-CXR-T supervisedStandard avg: 65.2 -> 64.1Hurt
replace CLIP loss with SigLIP loss when retraining (ablation)BioViL-T · MS-CXR-T supervisedStandard avg: 60.1 -> 61.1No clear effect
use paired images instead of single imageBioViL-T · MIMIC retrievalTEM: 3.5 -> 12.1Improved
add Change-aware Sigmoid lossBioViL-T SigLIP · MIMIC retrievalRecall@k / TEM: TEM 15.7 -> 17.1; I2T@5 35.1 -> 34.8Mixed
add TILA to BioViL-T SigLIP in zero-shotBioViL-T SigLIP · MS-CXR-T zero-shotConsistency avg: 24.0 -> 30.5Improved
add TILA to BioViL-T SigLIP in supervised training on private cohortBioViL-T SigLIP · Private cohortStandard avg: 55.3 -> 66.1Improved
use TILA representations for binary interval-change screeningBioViL-T · RexGradientAUC: 0.565 -> 0.702Improved

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