AI Paper Atlas

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

Revisiting Realistic Test-Time Training: Sequential Inference and Adaptation by Anchored Clustering

Yongyi Su, Xun Xu, Kui Jia

NeurIPS 2022 · Label-efficient and robust learning › Test-time adaptation · approach: Realistic and stable test-time adaptation

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 anchored clustering alone (all test samples update cluster statistics) (ablation)TTAC N-O baseline · CIFAR10-Cerror rate (%): 29.15 -> 14.32Improved
add pseudo label filtering alone (self-training on filtered labels) (ablation)no-adaptation baseline · CIFAR10-Cerror rate (%): 29.15 -> 15.00Improved
combine pseudo label filtering with anchored clustering (ablation)anchored clustering · CIFAR10-Cerror rate (%): 14.32 -> 11.33Improved
add global feature alignment with KL-Divergence on top of anchored clustering and filtering (ablation)TTAC N-O · CIFAR10-Cerror rate (%): 11.33 -> 10.94Improved
use KL-Divergence instead of L2 distance for global feature alignment (ablation)TTAC N-M · CIFAR10-Cerror rate (%): 11.87 -> 10.8Improved
replace L2 alignment with KL-Divergence in the TTT++ code (ablation)TTT++ (Y-M) · CIFAR10 level 5 snowerror rate (%): 9.85 -> 8.43Improved
add contrastive branch (modify source training) to TTAC (ablation)TTAC · CIFAR10-Cerror rate (%): 10.94 -> 10.69Improved
maintain a 4096-sample test queue instead of only the current batchTTAC · CIFAR10-Cerror rate (%): 11.91 -> 10.88Improved
maintain a test queue instead of only the current batchSHOT · CIFAR10-Cerror rate (%): 15.18 -> 13.96Improved
increase update epochs on the queue from 1 to 4TTAC · CIFAR10-Cerror rate (%): 10.88 -> 9.96Improved
increase update epochs from 4 to 6TTAC · CIFAR10-Caverage error rate (%): 10.94 -> 10.98No clear effect
update cluster statistics with all pseudo-labeled samples (no filtering) (ablation)TTAC · CIFAR10-Caverage error rate (%): 10.94 -> 13.00Hurt

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