Bidirectional cross-modality unsupervised domain adaptation using generative adversarial networks for cardiac image segmentation

Computer Science

• Generative Adversarial Networks 100%
• Image Segmentation 100%
• Unsupervised Domain Adaptation 100%
• Domain Adaptation 40%
• Performance Degradation 20%
• Image Alignment 20%
• Extracted Feature 20%
• image feature 20%
• Feature Map 20%
• Self-Attention Mechanism 20%
• Multimodality 20%
• Segmentation Task 20%
• Segmentation Performance 20%
• Evaluation Result 20%
• Knowledge Distillation 20%
• Network Segmentation 20%