Hierarchical Image Transformation and Multi-Level Features for Anomaly Defect Detection

Keyphrases

• Adaptive Learning System 33%
• Anomaly Detection 33%
• Application Data 33%
• Baseline Performance 33%
• Deep Learning Model 33%
• Defect Detection 100%
• Detection Network 33%
• Feature Exploration 33%
• Feature Module 33%
• Gaussian Mixture Model 33%
• Hierarchical Process 33%
• Hierarchical Transformation 33%
• High-dimensional Features 33%
• Idealized Images 66%
• Image Transformation 100%
• Industrial Datasets 33%
• Industrial Metals 33%
• Industrial Production Lines 33%
• Multi-level Features 100%
• Non-ideal Conditions 33%
• Non-ideality 66%
• Normal Behavior 33%
• Normality 33%
• Perturbance 33%
• Real-world Application 33%
• Small Size Sample 33%
• Training Data 33%
• Training Samples 33%
• Video Anomaly Detection 33%
• Work Environment 33%

Engineering

• Anomaly Detection 66%
• Deep Learning 33%
• Defect Detection 100%
• Experimental Result 33%
• Extracted Feature 33%
• Gaussian Mixture Model 33%
• Ideal Image 66%
• Image Transformation 100%
• Industrial Production 33%
• Level Feature 100%
• Production Line 33%
• Real World Application 33%

Computer Science

• Adaptive Learning 33%
• Anomaly Detection 66%
• Application Data 33%
• Convolutional Neural Network 33%
• Deep Learning 33%
• Dimensional Feature 33%
• Experimental Result 33%
• Extracted Feature 33%
• Gaussian Mixture Model 33%
• Image Transformation 100%
• Industrial Production 33%
• Normal Behavior 33%
• Preprocessing 33%
• Production Line 33%
• Training Data 33%
• Training Sample 33%
• World Application 33%