Research map / Label-efficient and robust learning / Test-time adaptation
Distill Once, Adapt Life-Long: Exploring Dataset Distillation for Continual Test-Time Adaptation
Hyun-Kurl Jang, Jihun Kim, Hyeokjun Kweon, KUK-JIN YOON
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 DO-ALL | EATA · ImageNet-C | avg error (%): 58.0 -> 56.6 | Improved |
| add DO-ALL | RMT · ImageNet-C | avg error (%): 59.8 -> 57.4 | Improved |
| add DO-ALL | ROID · ImageNet-C | avg error (%): 54.5 -> 53.6 | Improved |
| add DO-ALL | ASR · ImageNet-C | avg error (%): 54.3 -> 53.7 | Improved |
| add DO-ALL | ROID · CCC (Hard) | accuracy (%): 13.2 -> 15.5 | Improved |
| add DO-ALL | ROID · CCC (average) | accuracy (%): 33.6 -> 34.7 | Improved |
| add harm-adaptive blending (ablation) | DO-ALL (Replay+MMD+MixUp) · ImageNet-C (ROID) | avg error (%): 54.0 -> 53.6 | Improved |
| combine Replay, MMD and MixUp (ablation) | ROID · ImageNet-C | avg error (%): 54.5 -> 54.0 | Improved |
| use Target-Anchor MixUp instead of no MixUp (ablation) | DO-ALL · ImageNet-C (ROID) | avg error (%): 53.77 -> 53.55 | Improved |
| use Target-Target MixUp instead of Target-Anchor (ablation) | DO-ALL · ImageNet-C (ROID) | avg error (%): 53.55 -> 53.68 | Hurt |
| assign nearest anchor instead of farthest (ablation) | RMT + DO-ALL · ImageNet-C | avg error (%): 61.9 -> 57.4 | Improved |
| use SRe2L, DELT or WMDD for anchor distillation (ablation) | ROID + DO-ALL · ImageNet-C | avg error (%): 53.7 / 53.6 / 53.6 | 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 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
- ReservoirTTA: Prolonged Test-time Adaptation for Evolving and Recurring Domains (NeurIPS 2025 · similarity 0.90)
- Decorate the Newcomers: Visual Domain Prompt for Continual Test Time Adaptation (AAAI 2023 · similarity 0.89)
- DPCore: Dynamic Prompt Coreset for Continual Test-Time Adaptation (ICML 2025 · similarity 0.89)
- ViDA: Homeostatic Visual Domain Adapter for Continual Test Time Adaptation (ICLR 2024 · similarity 0.89)
- Towards Real-World Test-Time Adaptation: Tri-net Self-Training with Balanced Normalization (AAAI 2024 · similarity 0.88)
- A Versatile Framework for Continual Test-Time Domain Adaptation: Balancing Discriminability and Generalizability (CVPR 2024 · 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.
