AI Paper Atlas

Research map / Video / Video segmentation and tracking

Thermo-JEPA: Learning a Geometry-Grounded Thermal World Model via Cross-Modal Privileged Masking

Biwen Yang, Jin Zhang, Zhe Cao, Ruiheng Zhang

ECCV 2026 · Video › Video segmentation and tracking · approach: Video restoration with temporal priors · 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
add RGB-derived confidence-gated topology prior to the teacher attention (ablation)V-JEPA2 on thermal video · RGBT-World zero-shot action anticipationTop-1 accuracy (%): 18.6 -> 45.8Improved
add confidence gating to the structural prior (ablation)Thermo-JEPA · RGBT-World ablationTop-1 accuracy (%): 38.5 -> 45.8Improved
add stop-gradient on teacher targets (ablation)Thermo-JEPA · RGBT-World ablationTop-1 accuracy (%): 31.2 -> 45.8Improved
use DINOv2 semantic topology instead of Sobel edge prior (ablation)Thermo-JEPA · RGBT-World ablationTop-1 accuracy (%): 28.7 -> 45.8Improved
use random topology prior instead of DINOv2 topology (ablation)Thermo-JEPA · RGBT-World ablationTop-1 accuracy (%): 18.6 -> 12.4Hurt
Thermo-JEPA prior vs vanilla V-JEPA2 on spatial probing (ablation)Thermo-JEPA · RGBT-World diagnostic probeSpatial mIoU: 32.1 -> 68.7Improved
Thermo-JEPA vs vanilla V-JEPA2 on trajectory forecasting (ablation)Thermo-JEPA · RGBT-World diagnostic probeTrajectory ADE: 28.5 -> 8.4Improved
Thermo-JEPA vs large generic video foundation models, IR-only probeThermo-JEPA (300M) · RGBT-World zero-shot action anticipationTop-1 accuracy (%): 33.5 -> 45.8Improved
dense contrastive alignment of IR to RGB instead of privileged priorV-JEPA2 pretraining · RGBT-World zero-shot action anticipationTop-1 accuracy (%): 18.6 -> 24.2 vs 45.8 for Thermo-JEPAMixed

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