AI Paper Atlas

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

SePiCo: Semantic-Guided Pixel Contrast for Domain Adaptive Semantic Segmentation

Binhui Xie, Shuang Li, Mingjia Li, Chi Harold Liu, Gao Ming Huang, Guoren Wang

TPAMI 2023 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Adaptation for semantic segmentation · 234 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
use distribution-aware pixel contrast (DistCL) instead of centroid-aware ProtoCLSePiCo · GTAV to CityscapesmIoU: 58.8 -> 61.0Improved
use DistCL instead of memory-bank BankCLSePiCo · GTAV to CityscapesmIoU: 59.8 -> 61.0Improved
apply SePiCo (DistCL) self-training with pixel contrast over source-only trainingDeepLab-V2 ResNet-101 · GTAV to CityscapesmIoU: +22.4% mIoUImproved
apply SePiCo (DistCL) over source-only trainingDeepLab-V2 ResNet-101 · SYNTHIA to Cityscapes (16 classes)mIoU: +24.6% mIoUImproved
use SePiCo (DistCL) one-stage training instead of multi-stage ProDADeepLab-V2 ResNet-101 · GTAV to CityscapesmIoU: +3.5% mIoUImproved
use DistCL instead of ProtoCL or BankCLSePiCo · SYNTHIA to Cityscapes (16 classes)mIoU: 56.8 -> 58.1Improved
use DistCL instead of ProtoCLSePiCo · Cityscapes to Dark ZurichmIoU: 42.6 -> 45.4Improved
replace DeepLab-V2 with DAFormer architecture and add DistCLSePiCo · GTAV to Cityscapes (DAFormer)mIoU: 68.3 -> 70.3Improved
add DistCL to DAFormerSePiCo · Cityscapes to Dark Zurich (DAFormer)mIoU: 48.5 -> 54.2Improved
add ProtoCL to DAFormerSePiCo (ProtoCL) · GTAV to Cityscapes (DAFormer)mIoU: 68.3 -> 68.5No clear effect
add DistCL to DAFormer, test on unseen night domainsSePiCo · Nighttime Driving (generalization from Dark Zurich)mIoU: 51.8 -> 56.9Improved
add DistCL to DAFormer, test on unseen night domainsSePiCo · BDD100k-night (generalization from Dark Zurich)mIoU: 33.9 -> 40.6Improved

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