Co-training for visual object recognition based on self-supervised models using a cross-entropy regularization

Keyphrases

• Automatic Recognition 16%
• CIFAR-100 16%
• Co-training 100%
• Complementary Perspective 16%
• Cross-entropy 100%
• Deep Learning Methods 33%
• Deep Neural Network 16%
• Entropy Regularization 100%
• Image Data 16%
• Intermediate Inputs 16%
• Labeled Data 33%
• MNIST 16%
• Model Merging 16%
• Recent Approaches 16%
• Recognition-based 100%
• Relative Improvement 16%
• Self-supervised Learning 16%
• Self-supervised Models 100%
• Self-supervised Neural Network 16%
• Semi-supervised Model 16%
• Training Model 33%
• Visual Object 33%
• Visual Object Recognition 100%

Computer Science

• Deep Learning 50%
• Deep Learning Technique 50%
• Deep Neural Network 50%
• Learning Approach 50%
• Neural Network 50%
• Object Recognition 100%
• Regularization 100%
• Self-Supervised Learning 50%
• Semisupervised Model 50%
• Training Model 100%