Comparing Different Deep Learning Architectures as Vision-Based Multi-Label Classifiers for Identification of Multiple Distresses on Asphalt Pavement

Keyphrases

• Asphalt Pavement 100%
• Deep Learning Architectures 100%
• Vision-based 100%
• Distress Identification 100%
• Multi-label Classifier 100%
• Multi-label Classification 75%
• Network Level 50%
• Video Survey 50%
• Classification Algorithms 25%
• Computing Resources 25%
• Best Model 25%
• Classification Basis 25%
• Whole Network 25%
• Bleeding 25%
• Image-based 25%
• Classification Model 25%
• Convolutional Neural Network 25%
• Laser Scanner 25%
• Resource Consumption 25%
• Convolutional Neural Network Architecture 25%
• Versatile Applications 25%
• F1 Score 25%
• Pothole 25%
• Pavement Management System 25%
• VGG16 25%
• Pavement Management 25%
• Squeeze-and-excitation Network (SENet) 25%
• Classification Network 25%
• ResNet50 25%
• Traffic Speed 25%
• Road Condition Evaluation 25%
• Road Condition 25%
• Lightweight Classifier 25%
• Right-of-way 25%
• Network-level Pavement Management 25%

Engineering

• Asphalt Pavements 100%
• Deep Learning Method 100%
• Network Level 100%
• Road 66%
• Pavement 66%
• Convolutional Neural Network 66%
• Integral Part 33%
• Resource Consumption 33%
• Classification Algorithm 33%
• Laser Scanner 33%
• Neural Network Architecture 33%
• Classification Network 33%
• Right-of-Way 33%

Computer Science

• Deep Learning Method 100%
• Multi-Label Classification 100%
• Residual Neural Network 66%
• Convolutional Neural Network 66%
• Computer Hardware 33%
• Classification Algorithm 33%
• Neural Network Architecture 33%
• Classification Models 33%
• Potential Solution 33%
• Computing Resource 33%
• Resource Consumption 33%
• Average Accuracy 33%
• VGG16 33%