Research map / Label-efficient and robust learning / Test-time adaptation
Dual-level Adaptation for Multi-Object Tracking: Building Test-Time Calibration from Experience and Intuition
Wen Guo, Pengfei Zhao, Zongmeng Wang, Yufan Hu, Junyu Gao
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 TCEI test-time calibration (Intuitive plus Experiential) | MOTIP · DanceTrack test | HOTA: 69.5 -> 70.6 | Improved |
| add TCEI test-time calibration | MOTIP · SportsMOT test | HOTA: 72.6 -> 73.0 | Improved |
| add the Intuitive system (transient memory) (ablation) | TCEI · DanceTrack | HOTA: 69.5 -> 70.5 | Improved |
| add the Experiential system (experience cache) (ablation) | TCEI · DanceTrack | HOTA: 69.5 -> 70.4 | Improved |
| add the Experiential system on top of the Intuitive system (ablation) | TCEI · DanceTrack | HOTA: 70.5 -> 70.6 | Improved |
| use confident objects as temporal priors only (ablation) | TCEI · DanceTrack | HOTA: 69.5 -> 69.6 | No clear effect |
| use uncertain objects as reflective cases only (ablation) | TCEI · DanceTrack | HOTA: 69.5 -> 70.2 | Improved |
| combine confident and uncertain objects (ablation) | TCEI · DanceTrack | HOTA: 70.2 -> 70.5 | Improved |
| replace naive averaging with selective uncertainty-based calibration (ablation) | TCEI · DanceTrack | HOTA: 69.7 -> 70.6 | Improved |
| replace entropy-based selection with selective calibration (ablation) | TCEI · DanceTrack | HOTA: 69.9 -> 70.6 | Improved |
| apply Tent entropy-minimization backpropagation | MOTIP · DanceTrack test | HOTA: 69.5 -> 69.4 | No clear effect |
| apply Tent instead of TCEI on SportsMOT | MOTIP · SportsMOT test | HOTA: 72.6 -> 72.8 | No clear effect |
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.
- Tent — Tent: Fully Test-Time Adaptation by Entropy Minimization (ICLR 2021) · better on 5 of 5 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
- DARTH: Holistic Test-time Adaptation for Multiple Object Tracking (ICCV 2023 · similarity 0.89)
- DiffusionTrack: Diffusion Model for Multi-Object Tracking (AAAI 2024 · similarity 0.87)
- Multiple Object Tracking as ID Prediction (CVPR 2025 · similarity 0.87)
- Focusing on Tracks for Online Multi-Object Tracking (CVPR 2025 · similarity 0.87)
- UCMCTrack: Multi-Object Tracking with Uniform Camera Motion Compensation (AAAI 2024 · similarity 0.86)
- Sampling-Resilient Multi-Object Tracking (AAAI 2024 · similarity 0.86)
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.
