AI Paper Atlas

Research map / Continual learning / Domain and prompt-based

Reshaping the Online Data Buffering and Organizing Mechanism for Continual Test-Time Adaptation

Zhilin Zhu, Xiaopeng Hong, Zhiheng Ma, Weijun Zhuang, YaoHui Ma, Yong Dai, Yaowei Wang

ECCV 2024 · Continual learning › Domain and prompt-based · approach: Continual test-time adaptation · 9 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
replace reservoir buffering with uncertainty-aware buffering (with L_PCE) (ablation)OBAO · CIFAR100-to-CIFAR100Cmean error (%): 30.5 -> 29.7Improved
use FIFO buffering instead of uncertainty-aware buffering (ablation)OBAO · CIFAR100-to-CIFAR100Cmean error (%): 30.7 -> 30.6No clear effect
add class relation preservation constraint L_CRP alone (ablation)self-training baseline · CIFAR100-to-CIFAR100Cmean error (%): 30.7 -> 30.0Improved
add L_CRP on top of uncertainty buffering with replay (ablation)OBAO · CIFAR100-to-CIFAR100Cmean error (%): 29.7 -> 29.0Improved
add uncertainty-aware buffer with pseudo-target replay (ablation)self-training baseline · CIFAR100-to-CIFAR100Cmean error (%): 30.7 -> 29.7Improved
increase lambda_CRP over a broad range, but set 300 (ablation)OBAO · CIFAR100-to-CIFAR100Cmean error (%)Mixed
use OBAO instead of CoTTA on long-horizon segmentationSegformer-B5 · Cityscapes-to-ACDC (10 rounds)mean mIoU (%): 58.6 -> 61.3Improved
use entropy-minimizing TENT-cont. over long sequencesSegformer-B5 · Cityscapes-to-ACDC (10 rounds)mIoU (%): 56.7 -> 47.8 by round 10Hurt
use OBAO instead of DSS on ImageNet-CResNet-50 · ImageNet-to-ImageNet-Cmean error (%): 64.6 -> 59.0Improved
TENT-cont. on CIFAR100-to-CIFAR100C long sequenceResNeXt-29 · CIFAR100-to-CIFAR100Cmean error (%): 35.4 (BN adapt) -> 60.9Hurt

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