Pattern Collections & Testing
Pattern Collections
Adversarial glitch patterns designed to maximize facial detection confusion in surveillance algorithms.
Featured pattern test below:
Featured pattern test below:
AI Detection Test Results
- False face detections: 2
- False eye detections: 0
- False smile detections: 48
Pattern Analysis
- Edge density: 0.1463 (71,422 edge pixels)
- Contrast level: 86.27
- High frequency energy: 97,635
Adversarial Effectiveness
- False positive score: 50
- Complexity score: 120.89
- Overall adversarial score: 170.89
✅ EXCELLENT: High false positive rate - very confusing to AI
✅ HIGH COMPLEXITY: Dense edge patterns good for disruption
✅ HIGH COMPLEXITY: Dense edge patterns good for disruption
Anti-surveillance clothing paying homage to Palestine.
Featured pattern test below:
Featured pattern test below:
AI Detection Test Results
- False face detections: 1
- False eye detections: 0
- False smile detections: 200
Pattern Analysis
- Edge density: 0.1070 (452,362 edge pixels)
- Contrast level: 84.71
- High frequency energy: 845,415
Adversarial Effectiveness
- False positive score: 201
- Complexity score: 864.59
- Overall adversarial score: 1065.59
✅ EXCELLENT: High false positive rate - very confusing to AI
✅ HIGH COMPLEXITY: Dense edge patterns good for disruption
✅ HIGH COMPLEXITY: Dense edge patterns good for disruption
