Combating Ambiguity for Hash-Code Learning in Medical Instance Retrieval

Keyphrases

• Hash Function Learning 100%
• Instance Retrieval 100%
• Y-Net 100%
• Abnormal Region 75%
• Visual Features 50%
• Convolutional Features 50%
• Hash Code 50%
• Segmentation Loss 50%
• Classification Loss 50%
• State-of-the-art Techniques 25%
• Feature Extracting 25%
• Training Model 25%
• Radiologists 25%
• Medical Image Data 25%
• Semantic Information 25%
• Retrieval Practice 25%
• Evidence-based Diagnosis 25%
• Discriminative Features 25%
• Superpixel Segmentation 25%
• Similar Cases 25%
• Feature Aggregation 25%
• Segmentation Classification 25%
• Separability 25%
• Code Generating 25%
• Large Image Database 25%

Computer Science

• Visual Feature 100%
• Training Model 50%
• Image Database 50%
• retrieval performance 50%
• Discriminability 50%
• Discriminative Feature 50%