Long-Tailed Visual Recognition via Self-Heterogeneous Integration with Knowledge Excavation

Keyphrases

• Mixture-of-experts 100%
• Long-tailed Recognition 100%
• Heterogeneous Integration 100%
• Long-tailed Distribution 66%
• Knowledge Fusion 66%
• State-of-the-art Techniques 33%
• Classification Accuracy 33%
• Knowledge Transfer 33%
• Deep Model 33%
• Source Code 33%
• Majority Class 33%
• Deep Neural Network 33%
• ImageNet 33%
• Fused Features 33%
• Vanilla 33%
• Art Performance 33%
• Long-tailed Data 33%
• INaturalist 33%
• Hard Negative 33%
• Expert-based 33%
• CIFAR-100 33%
• Problem State 33%
• Model Depth 33%
• Dynamic Knowledge 33%

Computer Science

• Long-Tailed Visual Recognition 100%
• Heterogeneous Integration 100%
• Classification Accuracy 33%
• Majority Class 33%
• Deep Neural Network 33%
• Knowledge Transfer 33%
• Art Performance 33%

Engineering

• State-of-the-Art Method 100%
• Classification Accuracy 100%
• Deep Neural Network 100%