Research map / Vision-language and multimodal / Cross-modal learning
XTrack: Multimodal Training Boosts RGB-X Video Object Trackers
Yuedong Tan, Zongwei Wu, Yuqian Fu, Zhuyun Zhou, Guolei Sun, Eduard Zamfir, Chao Ma, Danda Paudel, Luc Van Gool, Radu Timofte
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 shared expert (ablation) | XTrack-B MeME · DepthTrack | F-score: 57.6 -> 61.5 | Improved |
| add modality-specific experts (ablation) | XTrack-B MeME · DepthTrack | F-score: 59.1 -> 61.5 | Improved |
| add shared expert (ablation) | XTrack-B MeME · VisEvent | Precision: 76.2 -> 77.5 | Improved |
| add modality-specific experts (ablation) | XTrack-B MeME · LasHeR | Precision: 67.8 -> 69.1 | Mixed |
| use two experts per modality instead of one (ablation) | XTrack-B · DepthTrack | F-score: 60.0 -> 61.5 | Improved |
| use two experts instead of three (ablation) | XTrack-B · DepthTrack | F-score: 60.9 -> 61.5 | Improved |
| train on event, depth and thermal instead of event only (ablation) | XTrack-B · DepthTrack | F-score: 45.5 -> 61.5 | Improved |
| train on event, depth and thermal instead of event only (ablation) | XTrack-B · LasHeR | Precision: 58.1 -> 69.1 | Improved |
| add depth data to event training (ablation) | XTrack-B · DepthTrack | F-score: 45.5 -> 60.8 | Improved |
| add depth data to event training (ablation) | XTrack-B · LasHeR | Precision: 58.1 -> 58.1 | No clear effect |
| use soft router (80%) instead of rigid classification (100%) (ablation) | XTrack-B · DepthTrack | F-score: 58.8 -> 61.5 | Improved |
| use 80% soft router instead of 33% near random assignment (ablation) | XTrack-B · VisEvent | Precision: 77.0 -> 77.5 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- SDSTrack — SDSTrack: Self-Distillation Symmetric Adapter Learning for Multi-Modal Visual Object Tracking (CVPR 2024) · better on 6 of 6 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Single-Model and Any-Modality for Video Object Tracking (CVPR 2024 · similarity 0.91)
- SDSTrack: Self-Distillation Symmetric Adapter Learning for Multi-Modal Visual Object Tracking (CVPR 2024 · similarity 0.90)
- Exploiting Multimodal Spatial-temporal Patterns for Video Object Tracking (AAAI 2025 · similarity 0.90)
- Collaborating Vision, Depth, and Thermal Signals for Multi-Modal Tracking: Dataset and Algorithm (NeurIPS 2025 · similarity 0.89)
- Bi-directional Adapter for Multimodal Tracking (AAAI 2024 · similarity 0.88)
- CSTrack: Enhancing RGB-X Tracking via Compact Spatiotemporal Features (ICML 2025 · similarity 0.88)
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.
