Prefallkd: Pre-Impact Fall Detection Via CNN-ViT Knowledge Distillation

Keyphrases

• Knowledge Distillation 100%
• Pre-impact Fall Detection 100%
• Aging Society 66%
• Student Model 66%
• Trained Teachers 33%
• Computational Complexity 33%
• Experiment Results 33%
• Detection Performance 33%
• Model Complexity 33%
• Mobile Devices 33%
• Deep Learning 33%
• Performance Complexity 33%
• Resource-constrained 33%
• Neural Network Model 33%
• State-of-the-art Models 33%
• Public Dataset 33%
• Latency Constraint 33%
• Imbalanced Data 33%
• Fall-related Injuries 33%
• Fall Detection System 33%
• Data Augmentation 33%
• Lightweight Neural Network 33%
• Simple Neural Networks 33%
• F1 Score 33%
• Severe Injury 33%
• Teacher Model 33%
• Vision Transformer 33%
• Fall Protection System 33%
• Model Lightweight 33%

Engineering

• Fall Detection 100%
• Convolutional Neural Network 100%
• Computational Complexity 33%
• Detection Performance 33%
• Lead Time 33%
• Network Model 33%
• Complex Model 33%
• Fall Accident 33%
• System Protection 33%
• Fall Protection 33%
• Deep Learning Method 33%

Computer Science

• Convolutional Neural Network 100%
• Knowledge Distillation 100%
• Computational Complexity 33%
• Neural Network Model 33%
• Detection Performance 33%
• Data Augmentation 33%
• Trained Teacher 33%
• Vision Transformer 33%
• Deep Learning Method 33%
• Mobile Device 33%