AI Paper Atlas

Research map / Label-efficient and robust learning / Semi-supervised learning

ATSO: Asynchronous Teacher-Student Optimization for Semi-Supervised Image Segmentation

Xinyue Huo, Lingxi Xie, Jianzhong He, Zijie Yang, Wengang Zhou, Houqiang Li, Qi Tian

CVPR 2021 · Label-efficient and robust learning › Semi-supervised learning · approach: Consistency training for semi-supervised segmentation · 68 citations (OpenAlex)

Check 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
Train each generation from scratch instead of initializing it from the previous teacher (ablation)Teacher-student optimization on NIH pancreas · NIH pancreas, 10% labeled, test setDSC (%): 78.98 -> 79.67Improved
Use asynchronous reference subsets so a teacher does not label data used to train its student (ablation)STSO on NIH pancreas · NIH pancreas, 10% labeled, test setDSC (%): 79.67 -> 81.69Improved
Use asynchronous reference subsets so a teacher does not label data used to train its student (ablation)STSO on NIH pancreas · NIH pancreas, 20% labeled, test setDSC (%): 83.21 -> 83.70Improved
Use asynchronous reference subsets so a teacher does not label data used to train its student (ablation)STSO on NIH-to-MSD pancreas transfer · MSD pancreas test setDSC (%): 75.48 -> 76.71Improved
Use asynchronous reference subsets rather than synchronous optimization (ablation)Teacher-student transfer from Cityscapes to Mapillary · Mapillary test, 19 classesmIoU (%): 28.11 -> 28.26No clear effect
Reduce pseudo-label classes from 19 to 5 (ablation)ATSO transfer from Cityscapes to Mapillary · Mapillary test, 5 classesmIoU (%): 69.94 -> 77.11Improved
Use 5-class pseudo labels from ATSO rather than 5-class pseudo labels from STSO (ablation)Target-network training on Mapillary · Mapillary test, 5-class pseudo-label supervisionmIoU (%): 29.76 -> 34.36Improved
Tune the 5-class ATSO model using 19-class pseudo labels (ablation)ATSO transfer from Cityscapes to Mapillary · Mapillary test, 19 classesmIoU (%): 34.36 -> 37.23Improved
Apply self-learning with the previous generation as initialization instead of training from scratch (ablation)NIH pancreas semi-supervised learning · NIH pancreas, 10% labeled, test setDSC (%): 78.98 at G2; 77.78 at G4Hurt
Use asynchronous optimization instead of synchronous optimization (ablation)Teacher-student optimization on Cityscapes · Cityscapes, 1K labeled, test setmIoU (%): 68.72 -> 70.43Improved

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