AI Paper Atlas

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

Rectifying Pseudo Label Learning via Uncertainty Estimation for Domain Adaptive Semantic Segmentation

Zhedong Zheng, Yi Yang

IJCV 2021 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Adaptation for semantic segmentation · 562 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 variance regularization (exp(-KL) weighted CE + KL) to pseudo label learningMRNet pseudo label learning · GTA5 -> CityscapesmIoU: 48.3 -> 50.3Improved
add variance regularization to pseudo label learningMRNet pseudo label learning · SYNTHIA -> Cityscapes (16 classes)mIoU: 46.5 -> 47.9Improved
add variance regularization to pseudo label learningMRNet pseudo label learning · SYNTHIA -> Cityscapes (13 classes)mIoU*: 53.8 -> 54.9Improved
add variance regularization to pseudo label learningMRNet pseudo label learning · Cityscapes -> Oxford RobotCarmIoU: 73.9 -> 74.4Improved
replace hand-crafted confidence threshold with variance regularization (ablation)pseudo label learning · GTA5 -> CityscapesmIoU: 48.4 -> 50.3Improved
raise confidence threshold to >0.99 (ablation)pseudo label learning · GTA5 -> CityscapesmIoU: 48.3 -> 45.5Hurt
use a moderate threshold >0.90 versus none (ablation)pseudo label learning · GTA5 -> CityscapesmIoU: 48.3 -> 48.4Improved
use pseudo labels from weaker AdaptSegNet with variance regularization (ablation)pseudo label learning · GTA5 -> Cityscapes (AdaptSegNet pseudo labels)mIoU: 46.8 -> 47.4Improved
add dropout (rate 0.1) in classifiers (ablation)variance regularization · GTA5 -> CityscapesmIoU: 49.6 -> 50.3Improved
vary dropout rate from 0.1 to 0.7 (ablation)variance regularization · GTA5 -> CityscapesmIoU: 50.3 -> 50.0No clear effect
use KL(F||F_aux) instead of MSE distance (ablation)variance regularization · GTA5 -> CityscapesmIoU: 49.6 -> 50.3Improved
use KL(F||F_aux) instead of swapped KL(F_aux||F) (ablation)variance regularization · GTA5 -> CityscapesmIoU: 49.4 -> 50.3Improved

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