Domain Adaptation On Gta5 To Cityscapes

评估指标

mIoU

评测结果

各个模型在此基准测试上的表现结果

Paper TitleRepository
HALO77.8Hyperbolic Active Learning for Semantic Segmentation under Domain Shift
DCF77.7Transferring to Real-World Layouts: A Depth-aware Framework for Scene Adaptation
ILM-ASSL76.1Iterative Loop Method Combining Active and Semi-Supervised Learning for Domain Adaptive Semantic Segmentation
MIC75.9MIC: Masked Image Consistency for Context-Enhanced Domain Adaptation
HRDA+PiPa75.6PiPa: Pixel- and Patch-wise Self-supervised Learning for Domain Adaptative Semantic Segmentation
HRDA73.8HRDA: Context-Aware High-Resolution Domain-Adaptive Semantic Segmentation
FREDOM - Transformer73.6FREDOM: Fairness Domain Adaptation Approach to Semantic Scene Understanding
SePiCo70.3SePiCo: Semantic-Guided Pixel Contrast for Domain Adaptive Semantic Segmentation
DAFormer + ProCST69.4ProCST: Boosting Semantic Segmentation Using Progressive Cyclic Style-Transfer
DAFormer68.3DAFormer: Improving Network Architectures and Training Strategies for Domain-Adaptive Semantic Segmentation
MIC + Guidance Training67.0Improve Cross-domain Mixed Sampling with Guidance Training for Adaptive Segmentation
TransDA-B63.9Smoothing Matters: Momentum Transformer for Domain Adaptive Semantic Segmentation
DDB62.7Deliberated Domain Bridging for Domain Adaptive Semantic Segmentation
EHTDI*62.0Exploring High-quality Target Domain Information for Unsupervised Domain Adaptive Semantic Segmentation
FREDOM - DeepLabV261.3FREDOM: Fairness Domain Adaptation Approach to Semantic Scene Understanding
BiSMAP61.2Bidirectional Self-Training with Multiple Anisotropic Prototypes for Domain Adaptive Semantic Segmentation
G2L59.7G2L: A Global to Local Alignment Method for Unsupervised Domain Adaptive Semantic Segmentation-
FAFS58.8A Novel Unsupervised Domain Adaption Method for Depth-Guided Semantic Segmentation Using Coarse-to-Fine Alignment-
ProDA+CRA58.6Cross-Region Domain Adaptation for Class-level Alignment-
ProDA57.5Prototypical Pseudo Label Denoising and Target Structure Learning for Domain Adaptive Semantic Segmentation
0 of 26 row(s) selected.
Domain Adaptation On Gta5 To Cityscapes | SOTA | HyperAI超神经