AI Paper Atlas

Research map / Label-efficient and robust learning / Domain adaptation and generalization

Let Synthetic Data Shine: Domain Reassembly and Soft-Fusion for Single Domain Generalization

Hao Li, Yubin Xiao, Ke Liang, MengZhu Wang, Long Lan, Kenli Li, X Y Liu

IJCV 2026 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Optimization-based domain generalization · 4 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 feature reassembly loss to baseline (ablation)DRSF · DWD SDG-OD (train Daytime-Clear)mPT: 25.3 -> 30.7Improved
add alignment loss to reassembly loss (ablation)DRSF · DWD SDG-OD (train Daytime-Clear)mPT: 30.7 -> 30.9Improved
add alignment loss to reassembly loss (ablation)DRSF · SDG-SS averagemIoU: 44.5 -> 45.7Improved
add adversarial soft fusion (MDSF) to DFDR (ablation)DRSF · DWD SDG-OD (train Daytime-Clear)mPT: 30.9 -> 32.2Improved
add DFDR pre-alignment to MDSF (ablation)DRSF · DWD SDG-OD (train Daytime-Clear)mPT: 29.2 -> 32.2Improved
add adversarial fusion to reassembly in SDG-SS (ablation)DRSF · SDG-SS averagemIoU: 45.7 -> 48.1Improved
add MIC UDA to DFDR+MDSF (ablation)DRSF · DWD SDG-OD (train Daytime-Clear)mPT: 32.2 -> 34.7Improved
apply DFDR to res1-3 instead of all four stages (ablation)DRSF · DWD SDG-OD (train Daytime-Clear) Night-RainymAP: 14.4 -> 20.2Improved
apply DFDR to res1 only (versus baseline) (ablation)DRSF · DWD SDG-OD (train Daytime-Clear) Night-RainymAP: 12.1 -> 18.7Improved
train on all five pseudo-target domains instead of one (ablation)DRSF · DWD SDG-OD (train Daytime-Clear) Dusk-RainymAP: 37.0 -> 39.8Improved
train on all five pseudo-domains instead of none (ablation)DRSF · DWD SDG-OD (train Daytime-Clear) sourcemAP: 50.2 -> 57.7Improved
rephrase foggy generation prompt (ablation)DRSF · DWD Daytime-FoggymAP: 38.6 -> 38.4No 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.

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