Handling Open-Set Noise and Novel Target Recognition in Domain Adaptive Semantic Segmentation

Keyphrases

• Target Recognition 100%
• Simt 100%
• Domain Adaptive Semantic Segmentation 100%
• Target Domain 25%
• Label Noise 25%
• Closed Set 25%
• Volume Regularization 25%
• Noise Distribution 25%
• Semantic Constraints 25%
• Segmentation Problem 12%
• Simplex 12%
• Adaptive Method 12%
• Domain Data 12%
• Two-component Domains 12%
• Generalization Ability 12%
• Regularizer 12%
• Geometric Properties 12%
• Segmentation Model 12%
• Domain Name 12%
• Black-box Model 12%
• Ground-truth Labels 12%
• Semantic Relations 12%
• Class Performance 12%
• Pseudo Label 12%
• Mixed Noise 12%
• Label Distribution 12%
• Domain Gap 12%
• Noise Transition Matrix 12%
• Label Correction 12%
• Domain Adaptive 12%
• Computational Geometry 12%
• Label Shift 12%
• Non-square 12%
• Geometry Analysis 12%
• Sample Recognition 12%
• Inter-class Relation 12%
• Simplex Noise 12%

Computer Science

• Image Segmentation 100%
• Target Recognition 100%
• Regularization 66%
• Simplex 66%
• Experimental Result 33%
• Transition Matrix 33%
• Generalization Ability 33%
• Target Domain Data 33%
• Segmentation Model 33%
• Semantic Relation 33%
• Label Distribution 33%
• Computational Geometry 33%
• Class Performance 33%

Engineering

• Target Recognition 100%
• Regularization 66%
• Simplex 66%
• Target Data 66%
• Noise Distribution 66%
• Experimental Result 33%
• Adaptive Method 33%
• Black-Box Model 33%