AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

Self-Supervised Vessel Segmentation via Adversarial Learning

Yuxin Ma, Yang Hua, Hanming Deng, Tao Song, Hao Wang, Zhengui Xue, Heng Cao, Ruhui Ma, Haibing Guan

ICCV 2021 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Self-supervised neuroimaging representations · 100 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 attention-guided generator (ablation)CycleGAN base with fractal maps · XCAD ablationJaccard: 0.229 -> 0.305Improved
add attention-guided generator (ablation)CycleGAN base with fractal maps · XCAD ablationSensitivity: 0.320 -> 0.487Improved
add segmentation loss (ablation)CycleGAN base with fractal maps · XCAD ablationJaccard: 0.229 -> 0.346Improved
add segmentation loss (ablation)CycleGAN base with fractal maps · XCAD ablationDice: 0.363 -> 0.511Improved
add segmentation loss on top of attention (ablation)Base + Attention · XCAD ablationJaccard: 0.305 -> 0.389Improved
add attention-guided generator on top of segmentation loss (ablation)Base + SegLoss · XCAD ablationJaccard: 0.346 -> 0.389Improved
combine attention and segmentation loss (ablation)Base · XCAD ablationAccuracy: 0.850 -> 0.945Improved
self-supervised training vs domain adaptation from DRIVEUNet/MMD/YNet · XCAD test (126 angiograms)Jaccard: 0.287 -> 0.389Improved
self-supervised vs unsupervised ReDOReDO · XCAD test (126 angiograms)Jaccard: 0.151 -> 0.389Improved
full supervision with annotations vs self-supervisedOurs · XCAD test (126 angiograms)Jaccard: 0.571 -> 0.389Hurt
self-supervised XCAD-pretrained model vs traditional on DRIVEKhan · DRIVEAUC: 0.885 -> 0.888No clear effect
self-supervised vs best supervised SoomroSoomro · DRIVEAUC: 0.948 -> 0.888Hurt

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