Research map / Detection and segmentation / Tracking
Spatio-Temporal Conditional Denoising Transformer for Modality-Missing RGBT Tracking
Andong Lu, Ziyi Zha, Jiandong Jin, Shihao Li, Chenglong Li, Jin Tang, Bin Luo
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add spatial condition to the denoiser (ablation) | SCDT baseline · LasHeR-Miss | PR: 63.2 -> 66.9 | Improved |
| add short-term temporal condition on top of spatial (ablation) | SCDT w/ SP · LasHeR-Miss | PR/SR: +1.4%/+1.4% | Improved |
| add long-term temporal condition on top of spatial and short-term (ablation) | SCDT w/ SP ST · LasHeR | PR/SR: +0.3%/+0.1% | Improved |
| combine short-term and long-term conditions (ablation) | SCDT best single-condition variant · LasHeR | PR/SR: +1.4% PR and +1.3% SR | Improved |
| combine short-term and long-term conditions (ablation) | SCDT best single-condition variant · LasHeR-Miss | PR/SR: +1.0%/+0.8% | Improved |
| use weak noise for complete and strong noise for missing instead of strong for both (ablation) | SCDT · LasHeR | PR: 73.2 -> 77.4 | Improved |
| use weak noise for complete and strong for missing instead of weak for both (ablation) | SCDT · LasHeR-Miss | PR: 68.9 -> 69.3 | Improved |
| add reconstruction loss to alignment loss (ablation) | SCDT · LasHeR-Miss | PR: 67.2 -> 69.3 | Improved |
| add alignment loss to reconstruction loss (ablation) | SCDT · LasHeR | PR: 75.6 -> 77.4 | Improved |
| use 4 denoising layers instead of 2 (ablation) | SCDT · LasHeR | PR: 75.9 -> 77.4 | Improved |
| use 6 denoising layers instead of 4 (ablation) | SCDT · LasHeR | PR: 77.4 -> 76.4 | Hurt |
| use 6 denoising layers instead of 4 (ablation) | SCDT · LasHeR-Miss | PR: 69.3 -> 69.2 | No clear effect |
Closest papers by meaning
- Temporal Adaptive RGBT Tracking with Modality Prompt (AAAI 2024 · similarity 0.88)
- Cross-modulated Attention Transformer for RGBT Tracking (AAAI 2025 · similarity 0.88)
- Single-Model and Any-Modality for Video Object Tracking (CVPR 2024 · similarity 0.87)
- RGBT Tracking via All-layer Multimodal Interactions with Progressive Fusion Mamba (AAAI 2025 · similarity 0.87)
- Collaborating Vision, Depth, and Thermal Signals for Multi-Modal Tracking: Dataset and Algorithm (NeurIPS 2025 · similarity 0.86)
- CSTrack: Enhancing RGB-X Tracking via Compact Spatiotemporal Features (ICML 2025 · similarity 0.85)
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.
