DEVOURED
Simon Weckert creates Digital Camouflage shirt to avoid AI video surveillance
Designer Simon Weckert created a patterned shirt designed to trigger misclassifications in computer vision systems, effectively camouflaging the wearer from AI surveillance.
What: Simon Weckert developed a piece of clothing featuring specific visual artifacts that exploit vulnerabilities in AI-driven object detection and tracking cameras.
Why it matters: This demonstrates the fragility of current computer vision models and how easily physical-world adversarial inputs can bypass safety and surveillance systems.
Decoder
- Adversarial Input: Data designed to trick machine learning models into making incorrect predictions.
Original article
Designer Simon Weckert created a colorful AI-generated camouflage shirt that exploits weaknesses in computer vision systems, making the wearer harder for AI surveillance cameras to recognize as a person.
Explore all 56 articles in today's digest
Fresh dev news, AI-digested daily. Never miss a trend at devoured.fyi