| UniMatch V2: Pushing the Limit of Semi-Supervised Semantic Segmentation | ✓ Link | 84.5% | UniMatch V2 (DINOv2-B) | 2024-10-14 |
| SemiVL: Semi-Supervised Semantic Segmentation with Vision-Language Guidance | ✓ Link | 80.3% | SemiVL (ViT-B/16) | 2023-11-27 |
| Revisiting and Maximizing Temporal Knowledge in Semi-supervised Semantic Segmentation | ✓ Link | 80.1% | PrevMatch (ResNet-101) | 2024-05-31 |
| FARCLUSS: Fuzzy Adaptive Rebalancing and Contrastive Uncertainty Learning for Semi-Supervised Semantic Segmentation | ✓ Link | 80.0 | FARCLUSS | 2025-06-11 |
| Semi-Supervised Semantic Segmentation via Marginal Contextual Information | ✓ Link | 79.52% | S4MC | 2023-08-26 |
| Switching Temporary Teachers for Semi-Supervised Semantic Segmentation | ✓ Link | 79.46 | Dual Teacher (DeepLab v3+ with ResNet-101 pretraind on ImageNet-1K) | 2023-09-21 |
| CorrMatch: Label Propagation via Correlation Matching for Semi-Supervised Semantic Segmentation | ✓ Link | 79.4% | CorrMatch (Deeplabv3+ with ResNet-101) | 2023-06-07 |
| Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic Segmentation | ✓ Link | 79.22% | UniMatch (DeepLab v3+ with ImageNet-pretrained ResNet-101, single scale inference) | 2022-08-21 |
| Semi-Supervised Semantic Segmentation with Cross Pseudo Supervision | ✓ Link | 79.21% | CPS (DeepLab v3+ with ImageNet-pretrained ResNet-101, single scale inference) | 2021-06-02 |
| Semi-Supervised Semantic Segmentation via Adaptive Equalization Learning | ✓ Link | 79.01% | AEL (DeepLab v3+ with ResNet-101 pretraind on ImageNet-1K) | 2021-10-11 |
| Revisiting and Maximizing Temporal Knowledge in Semi-supervised Semantic Segmentation | ✓ Link | 78.8% | PrevMatch (ResNet-50) | 2024-05-31 |
| Semi-Supervised Semantic Segmentation Using Unreliable Pseudo-Labels | ✓ Link | 78.51% | U2PL (DeepLab v3+ with ResNet-101 pretraind on ImageNet-1K, AEL) | 2022-03-08 |
| Confidence-Weighted Boundary-Aware Learning for Semi-Supervised Semantic Segmentation | ✓ Link | 78.43% | CW-BASS (DeepLab v3+ with ResNet-50) | 2025-02-21 |
| n-CPS: Generalising Cross Pseudo Supervision to n Networks for Semi-Supervised Semantic Segmentation | | 78.41% | n-CPS (ResNet-50) | 2021-12-14 |
| Semi-supervised Semantic Segmentation with Prototype-based Consistency Regularization | ✓ Link | 78.4% | PCR (DeepLab v3+ with ResNet-101 pretraind on ImageNet-1K) | 2022-10-10 |
| Perturbed and Strict Mean Teachers for Semi-supervised Semantic Segmentation | ✓ Link | 78.38% | PS-MT (DeepLab v3+ with ImageNet-pretrained ResNet-50, single scale inference) | 2021-11-25 |
| LaserMix for Semi-Supervised LiDAR Semantic Segmentation | ✓ Link | 78.3% | LaserMix (DeepLab v3+, ImageNet pre- trained ResNet50, single scale inference) | 2022-06-30 |
| A Simple Baseline for Semi-supervised Semantic Segmentation with Strong Data Augmentation | ✓ Link | 77.8% | SimpleBaseline(DeepLabv3+ with ImageNet pretrained Xception65, single scale inference) | 2021-04-15 |
| Conservative-Progressive Collaborative Learning for Semi-supervised Semantic Segmentation | ✓ Link | 76.98% | CPCL (DeepLab v3+ with ResNet-50) | 2022-11-30 |
| Semi-supervised Semantic Segmentation with Error Localization Network | ✓ Link | 73.52% | Error Localization Network (DeeplabV3 with ResNet-50) | 2022-04-05 |
| Three Ways to Improve Semantic Segmentation with Self-Supervised Depth Estimation | ✓ Link | 69.38% | SegSDE (MTL decoder with ResNet101, ImageNet pretrained, unlabeled image sequences) | 2020-12-19 |
| Bootstrapping Semantic Segmentation with Regional Contrast | ✓ Link | 68.50% | ReCo (DeepLab v3+ with ResNet-101 backbone, ImageNet pretrained) | 2021-04-09 |
| Bootstrapping Semantic Segmentation with Regional Contrast | ✓ Link | 67.53% | ReCo (DeepLab v2 with ResNet-101 backbone, ImageNet pretrained) | 2021-04-09 |
| GuidedMix-Net: Learning to Improve Pseudo Masks Using Labeled Images as Reference | ✓ Link | 67.5% | GuidedMix-Net(DeepLab v2 with ResNet101, ImageNet pretrained) | 2021-06-29 |
| Semi-Supervised Semantic Segmentation with Pixel-Level Contrastive Learning from a Class-wise Memory Bank | ✓ Link | 65.9% | SemiSegContrast
(DeepLab v2 with ResNet-101 backbone, MSCOCO pretrained) | 2021-04-27 |
| The GIST and RIST of Iterative Self-Training for Semi-Supervised Segmentation | | 65.14% | GIST and RIST (DeepLabv2 with ResNet101, MSCOCO pre-trained) | 2021-03-31 |
| Semi-supervised semantic segmentation needs strong, varied perturbations | ✓ Link | 63.87% | CutMix (DeepLab v2, ImageNet pre-trained) | 2019-06-05 |
| ClassMix: Segmentation-Based Data Augmentation for Semi-Supervised Learning | ✓ Link | 63.63% | ClassMix (DeepLab v2 MSCOCO pretrained) | 2020-07-15 |
| Semi-Supervised Semantic Segmentation with High- and Low-level Consistency | ✓ Link | 61.9% | s4GAN (DeepLab v2 ImageNet pre-trained) | 2019-08-15 |
| Adversarial Learning for Semi-Supervised Semantic Segmentation | ✓ Link | 60.5% | Adversarial (DeepLab v2 ImageNet pre-trained) | 2018-02-22 |