AI Paper Atlas

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

CODA: Generalizing to Open and Unseen Domains with Compaction and Disambiguation

Chaoqi Chen, Luyao Tang, Yue Huang, Xiaoguang Han, Yizhou Yu

NeurIPS 2023 · Label-efficient and robust learning › Test-time adaptation · approach: Characterizing out-of-distribution failures · 2 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 source compaction (no target disambiguation) (ablation)ERM · PACSH-score: 36.1 -> 51.2Improved
add source compaction (ablation)ERM · Office-HomeH-score: 34.3 -> 64.8Improved
add source compaction (ablation)ERM · DigitsH-score: 18.0 -> 39.0Improved
add target disambiguation only (ablation)ERM · PACSH-score: 36.1 -> 49.8Improved
add target disambiguation on top of source compaction (ablation)CODA · PACSH-score: 51.2 -> 58.3Improved
add target disambiguation on top of source compaction (ablation)CODA · Office-31H-score: 75.3 -> 81.0Improved
add target disambiguation on top of source compaction (ablation)CODA · DigitsH-score: 39.0 -> 41.3Improved
combine Eq. 3 and Eq. 4 instead of Eq. 3 alone (ablation)source compaction · PACSH-score: 46.8 -> 51.2Improved
combine Eq. 3 and Eq. 4 instead of Eq. 3 alone (ablation)source compaction · DigitsH-score: 38.6 -> 39.0Mixed
Eq. 2 objective instead of Eq. 1 (ablation)unknown-aware training · PACSH-score: 36.1 -> 41.1Improved
Eq. 3 virtual unknown objective instead of Eq. 1 (ablation)unknown-aware training · Office-31H-score: 40.7 -> 72.6Improved
TAST test-time adaptation during testingTAST · PACSaccuracy known vs unknownHurt

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