AI Paper Atlas

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

CADTrack: Learning Contextual Aggregation with Deformable Alignment for Robust RGBT Tracking

Hao Li, Yuhao Wang, Xiantao Hu, Wenning Hao, Pingping Zhang, Dong Wang, Huchuan Lu

AAAI 2026 · Vision-language and multimodal › Cross-modal learning · approach: Multi-modality image fusion · 26 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 deformable alignment module (DAM) (ablation)CADTrack baseline · LasHeRPR: 69.1 -> 72.7Improved
add Mamba-based feature interaction (MFI) (ablation)CADTrack baseline · LasHeRPR: 72.7 -> 74.1Improved
add contextual aggregation module (CAM) (ablation)CADTrack baseline · LasHeRPR: 74.1 -> 75.8Improved
add template update (ablation)baseline ViT-B tracker · LasHeRPR: 67.9 -> 69.1Improved
use DropMAE pretraining instead of SOT (ablation)CADTrack · LasHeRPR: 75.8 -> 77.7Improved
use MFI instead of TBSI for modality interaction (ablation)CADTrack · LasHeRPR: 74.5 -> 77.7 (FPS 29 -> 40)Improved
use MFI instead of BSI (ablation)CADTrack · LasHeRPR: 76.9 -> 77.7Improved
place MFI at layers 4, 7 and 10 instead of fewer positions (ablation)CADTrack · LasHeRPR: 75.2 -> 77.7Improved
use 2 Mamba layers in MFI instead of 1 (ablation)MFI · LasHeRPR: 75.4 -> 77.7Improved
use 3 Mamba layers in MFI instead of 2 (ablation)MFI · LasHeRPR: 77.7 -> 77.1Hurt
select top-6 experts versus 4 or 8 in CAM (ablation)CAM · LasHeRPR: k=4 75.5, k=6 77.7, k=8 74.3Mixed
use learned router selection instead of fixed-interval layer selection (ablation)CAM · LasHeRPR: 75.8 -> 77.7Improved

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