Semi Supervised Semantic Segmentation On 21

Metriken

Validation mIoU

Ergebnisse

Leistungsergebnisse verschiedener Modelle zu diesem Benchmark

		Paper Title
AllSpark	81.65	AllSpark: Reborn Labeled Features from Unlabeled in Transformer for Semi-Supervised Semantic Segmentation
PrevMatch (ResNet-101)	81.4	Revisiting and Maximizing Temporal Knowledge in Semi-supervised Semantic Segmentation
CorrMatch (Deeplabv3+ with ResNet-101)	81.3	CorrMatch: Label Propagation via Correlation Matching for Semi-Supervised Semantic Segmentation
UniMatch (DeepLab v3+ with ResNet-101)	80.94	Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic Segmentation
S4MC	78.84	Semi-Supervised Semantic Segmentation via Marginal Contextual Information
Dual Teacher	78.82	Switching Temporary Teachers for Semi-Supervised Semantic Segmentation
PCR (DeepLab v3+ with ResNet-101 pretraind on ImageNet-1K)	78.6	Semi-supervised Semantic Segmentation with Prototype-based Consistency Regularization
U2PL (DeepLab v3+ with ResNet-101 pretraind on ImageNet-1K, CutMix)	77.21	Semi-Supervised Semantic Segmentation Using Unreliable Pseudo-Labels
AEL (DeepLab v3+ with ResNet-101 pretraind on ImageNet-1K)	77.2	Semi-Supervised Semantic Segmentation via Adaptive Equalization Learning
n-CPS (ResNet-101)	75.86	n-CPS: Generalising Cross Pseudo Supervision to n Networks for Semi-Supervised Semantic Segmentation
CPCL (DeepLab v3+ with ResNet-101)	73.44	Conservative-Progressive Collaborative Learning for Semi-supervised Semantic Segmentation
n-CPS (ResNet-50)	72.03	n-CPS: Generalising Cross Pseudo Supervision to n Networks for Semi-Supervised Semantic Segmentation
CPCL (DeepLab v3+ with ResNet-50)	71.66	Conservative-Progressive Collaborative Learning for Semi-supervised Semantic Segmentation
PCT (DeepLab v3+ with ResNet-50 pretrained on ImageNet-1K)	71.35	Learning Pseudo Labels for Semi-and-Weakly Supervised Semantic Segmentation
Dense FixMatch (DeepLabv3+ ResNet-101, over-sampling, single pass eval)	54.85	Dense FixMatch: a simple semi-supervised learning method for pixel-wise prediction tasks
Dense FixMatch (DeepLabv3+ ResNet-50, over-sampling, single pass eval)	52.15	Dense FixMatch: a simple semi-supervised learning method for pixel-wise prediction tasks

0 of 16 row(s) selected.