AI Paper Atlas

Research map / Label-efficient and robust learning / Test-time adaptation

Improving Test-Time Adaptation via Shift-Agnostic Weight Regularization and Nearest Source Prototypes

Sungha Choi, Seunghan Yang, Seokeon Choi, Sungrack Yun

ECCV 2022 · Label-efficient and robust learning › Test-time adaptation · approach: Realistic and stable test-time adaptation · 62 citations (OpenAlex)

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 changedWhereMeasuredEffect
add SWR to main entropy loss at lr 1e-3 (ablation)Main (entropy) TTA · CIFAR-100-C, ResNet-50avg error (%): 75.70 -> 37.76Improved
add NSP auxiliary task to main loss (ablation)Main (entropy) TTA · CIFAR-100-C, ResNet-50avg error (%): 39.44 -> 37.55Improved
add NSP on top of SWR (ablation)Main + SWR · CIFAR-100-C, ResNet-50avg error (%): 37.76 -> 35.65Improved
add self-supervised aux loss to NSP entropy loss (ablation)Main + SWR + NSP · CIFAR-100-C, ResNet-50avg error (%): 35.80 -> 35.65Improved
attach projector (depth 2) before NSP (ablation)SWR+NSP · CIFAR-100-C, ResNet-50avg error (%): 36.34 -> 35.65Improved
attach projector (depth 2) before NSP (ablation)SWR+NSP · CIFAR-10-C, ResNet-50avg error (%): 12.52 -> 12.95Hurt
train projector with color, crop and blur transforms instead of none (ablation)NSP projector · CIFAR-100-C, ResNet-50avg error (%): 37.88 -> 35.65Improved
freeze projector during adaptation instead of fine-tuning (ablation)SWR+NSP · CIFAR-100-C, ResNet-50avg error (%): 35.96 -> 35.65Improved
update theta* each step instead of freezing it at source parameters (ablation)SWR · CIFAR-10-C, WRN-28-10avg error (%): 18.30 -> 15.70Improved
build prototypes from projection z instead of classifier weights (ablation)NSP · CIFAR-100-C, ResNet-50avg error (%): 39.89 -> 35.65Improved
raise SWR weight lambda_r from 10 to 250 (ablation)SWR+NSP · CIFAR-100-C, ResNet-50avg error (%): 53.01 -> 35.65Improved
raise SWR weight lambda_r from 250 to 1000 (ablation)SWR+NSP · CIFAR-100-C, ResNet-50avg error (%): 35.65 -> 37.43Hurt

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.

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

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy