Semi Supervised Semantic Segmentation On 8

评估指标

Validation mIoU

评测结果

各个模型在此基准测试上的表现结果

Paper TitleRepository
UniMatch V2 (DINOv2-B)85.1%UniMatch V2: Pushing the Limit of Semi-Supervised Semantic Segmentation
SemiVL (ViT-B/16)80.6%SemiVL: Semi-Supervised Semantic Segmentation with Vision-Language Guidance
Dual Teacher80.52Switching Temporary Teachers for Semi-Supervised Semantic Segmentation-
CorrMatch (Deeplabv3+ with ResNet-101)80.4%CorrMatch: Label Propagation via Correlation Matching for Semi-Supervised Semantic Segmentation
AEL (DeepLab v3+ with ResNet-101 pretraind on ImageNet-1K)80.28%Semi-Supervised Semantic Segmentation via Adaptive Equalization Learning
CPS (DeepLab v3+ with ImageNet-pretrained ResNet-101, single scale inference)80.21%Semi-Supervised Semantic Segmentation with Cross Pseudo Supervision
PrevMatch (ResNet-101)80.1%Revisiting and Maximizing Temporal Knowledge in Semi-supervised Semantic Segmentation
S4MC79.76%Semi-Supervised Semantic Segmentation via Marginal Contextual Information
UniMatch79.5%Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic Segmentation
n-CPS (ResNet-50)79.29%n-CPS: Generalising Cross Pseudo Supervision to n Networks for Semi-Supervised Semantic Segmentation-
PS-MT (DeepLab v3+ with ImageNet-pretrained ResNet-50, single scale inference)79.22%Perturbed and Strict Mean Teachers for Semi-supervised Semantic Segmentation
PrevMatch (ResNet-50)79.2%Revisiting and Maximizing Temporal Knowledge in Semi-supervised Semantic Segmentation
U2PL (DeepLab v3+ with ResNet-101 pretraind on ImageNet-1K, AEL)79.12%Semi-Supervised Semantic Segmentation Using Unreliable Pseudo-Labels
PCR (DeepLab v3+ with ResNet-101 pretraind on ImageNet-1K)79.11%Semi-supervised Semantic Segmentation with Prototype-based Consistency Regularization
LaserMix (DeepLab v3+, ImageNet pre- trained ResNet50, single scale inference)79.1%LaserMix for Semi-Supervised LiDAR Semantic Segmentation
SimpleBaseline(DeepLabv3+ with ImageNet pretrained Xception65, single scale inference)78.7%A Simple Baseline for Semi-supervised Semantic Segmentation with Strong Data Augmentation
CPCL (DeepLab v3+ with ResNet-50)78.17%Conservative-Progressive Collaborative Learning for Semi-supervised Semantic Segmentation
Error Localization Network (DeeplabV3 with ResNet-50)75.33%Semi-supervised Semantic Segmentation with Error Localization Network
GuidedMix-Net(DeepLab v2 with ResNet101, ImageNet pretrained)69.8%GuidedMix-Net: Learning to Improve Pseudo Masks Using Labeled Images as Reference
ReCo (DeepLab v2 with ResNet-101 backbone, ImageNet pretrained)68.69%Bootstrapping Semantic Segmentation with Regional Contrast
0 of 22 row(s) selected.
Semi Supervised Semantic Segmentation On 8 | SOTA | HyperAI超神经