Digital Camouflage: Wearable Shield Against AI Surveillance
Digital Camouflage, a conceptual clothing line by Simon Weckert, aims to interfere with AI-driven person detection technologies. Each piece is adorned with a continuous Adversarial Texture (AdvTexture) that covers its entirety, created through a unique generative AI technique known as TC-EGA. This design generates high-frequency visual noise and misleading visual elements from various perspectives, effectively tackling the segment-missing issue seen in earlier fixed-patch evasion methods. Produced in Latvia, the collection is made from a mix of 65% recycled polyester and 35% polyester, with the texture applied through digital textile printing. Weckert's project delves into themes of visibility, anonymity, and digital surveillance, questioning machine-based visual recognition. It was submitted to designboom's DIY submissions, with editing by Christina Vergopoulou.
Key facts
- Digital Camouflage is a conceptual garment collection by Simon Weckert.
- The garments are designed to interfere with AI systems that use computer vision to detect people.
- The pattern is based on the concept of an adversarial attack, manipulating visual data so machines misinterpret what they see.
- Previous attempts used fixed printed patches, which become ineffective due to the segment-missing problem.
- The collection uses a seamless Adversarial Texture (AdvTexture) covering the entire garment.
- The texture is generated using a specialized generative AI method called TC-EGA.
- The collection is manufactured in Latvia using a blend of 65% recycled polyester and 35% polyester.
- The adversarial texture is applied through digital textile printing.
- The project was submitted to designboom's DIY submissions feature.
- Edited by Christina Vergopoulou for designboom.
Entities
Artists
- Simon Weckert
Institutions
- designboom
Locations
- Latvia