Edge computing accelerated defect classification based on deep convolutional neural network with application in rolling image inspection

Mathematics

• Inspection 89%
• Defects 81%
• Neural Networks 76%
• Computing 50%
• Framework 46%
• Real-time 36%
• Saliency Map 32%
• Transfer Learning 30%
• Defect Detection 29%
• Learning 29%
• Requirements 26%
• Random Forest 23%
• Data Augmentation 21%
• Decision tree 20%
• High Speed 19%
• Learning Algorithm 19%
• Small Sample Size 18%
• Machine Learning 17%
• Visualization 17%
• Training 17%
• Prediction 13%
• Model 10%

Engineering & Materials Science

• Edge computing 100%
• Deep neural networks 78%
• Convolutional neural networks 74%
• Inspection 61%
• Defects 55%
• Deep learning 48%
• Transfer learning 32%
• Random forests 28%
• Decision trees 26%
• Learning algorithms 22%
• Visualization 20%
• Machine learning 19%

Chemical Compounds

• Time 29%
• Application 25%