Verification of Deep Convolutional Neural Networks Using ImageStars

Mathematics

• Attack 95%
• Neural Networks 88%
• Pose Estimation 39%
• Image Classification 38%
• Zonotope 34%
• Efficient Set 31%
• Scalability 30%
• Real-world Applications 29%
• Segmentation 28%
• Training 24%
• High Accuracy 24%
• Experimental Results 23%
• Polytope 21%
• Linear programming 21%
• Human 21%
• Semantics 20%
• Robustness 19%
• Verify 19%
• Evaluate 18%
• Output 17%
• Methodology 17%
• Framework 13%

Engineering & Materials Science

• Image classification 100%
• Deep neural networks 91%
• Linear programming 86%
• Convolutional neural networks 85%
• Scalability 75%
• Semantics 64%
• Concretes 63%
• Neural networks 56%