Research map / Label-efficient and robust learning / Test-time adaptation
Energy Waveify and Redistribution for Test-Time Adaptation: A Control System Perspective
Zhenbin Wang, Lei Zhang, Lituan Wang, Zhenwei Zhang, Guangwu Qian, Yan Wang, Wei Huang
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 |
|---|---|---|---|
| adapt with APT (wave-based energy redistribution) instead of Energy TEA | WRN-28-10 BatchNorm · CIFAR-10-C severity 5 | accuracy: 83.34 -> 86.93 | Improved |
| adapt with APT instead of Energy TEA | WRN-28-10 BatchNorm · CIFAR-100-C severity 1-5 | accuracy: 71.22 -> 73.57 | Improved |
| adapt with APT instead of Energy TEA | ResNet-50 GroupNorm · TinyImageNet-200-C severity 1-5 | accuracy: 30.41 -> 35.45 | Improved |
| adapt with APT instead of best baseline on average accuracy | WRN-28-10 BatchNorm · CIFAR/TinyImageNet-C average over severities | accuracy: +2.21% average over best baseline | Improved |
| adapt with APT instead of best baseline | ResNet-50 GroupNorm · CIFAR/TinyImageNet-C average over severities | accuracy: +2.30% average over best baseline | Improved |
| use sampling-free wave-based adaptation instead of top-3 baseline methods | APT · CIFAR-10-C / CIFAR-100-C, P100 GPU | adaptation time: 1/3 to 1/7 of the time of the top three baselines | Improved |
| adapt with APT instead of Energy TEA | WRN-28-10 BatchNorm · CIFAR-10 calibration | MCE: 47.37% -> 42.74% | Improved |
| adapt with entropy-based CRKD | WRN-28-10 BatchNorm · CIFAR-10 calibration | MCE: 57.99% -> 58.28% | Hurt |
| adapt with APT | PACS single-source · PACS all four source domains | accuracy: APT averages 29.13, 44.24, 35.01, 27.54 | Improved |
| increase adaptation steps (ablation) | APT · CIFAR-10-C, CIFAR-100-C, TinyImageNet-200-C | left/right energy ratio and accuracy | 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.
- TENT — Tent: Fully Test-Time Adaptation by Entropy Minimization (ICLR 2021) · better on 1 of 1 shared result
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
- Bayesian Weight Enhancement with Steady-State Adaptation for Test-time Adaptation in Dynamic Environments (ICML 2025 · similarity 0.84)
- TEA: Test-time Energy Adaptation (CVPR 2024 · similarity 0.84)
- Evaluation of Test-Time Adaptation Under Computational Time Constraints (ICML 2024 · similarity 0.84)
- Predicting the Energy Landscape of Stochastic Dynamical System via Physics-informed Self-supervised Learning (ICLR 2025 · similarity 0.84)
- Unified Entropy Optimization for Open-Set Test-Time Adaptation (CVPR 2024 · similarity 0.83)
- A Layer Selection Approach to Test Time Adaptation (AAAI 2025 · similarity 0.83)
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.
