Exploring Minimally Sufficient Representation in Active Learning through Label-Irrelevant Patch Augmentation

Computer Science

• Active Learning 100%
• Image Classification 28%
• Classification Task 14%
• Performance Model 14%
• Computational Cost 14%
• Self-Supervised Learning 14%
• Medical Imaging 14%
• Learned Feature 14%
• Learning Framework 14%
• Deep Learning Model 14%

Keyphrases

• Patch Augmentation 100%
• Self-supervised Learning Model 20%