Technology | Adversarial Fashion

Technology | Adversarial Fashion
  • Privacy in the Digital Age 👁



Our faces are a unique and intrinsic part of our identity, accompanying us throughout our lives. As the most personal data we possess, protecting our facial privacy is essential. This is where the concept of Privacy By Design (PBD) comes in, a principle enshrined in the General Data Protection Regulation (GDPR) of 2016. PBD emphasizes integrating privacy measures directly into the development of technology, ensuring that privacy concerns are addressed from the outset.



  • AI Camouflage | Confusing the Algorithms 📹



AI Camouflage is a new form of camouflage specifically designed for the AI era. This approach uses AI technologies to confuse AI algorithms, preventing them from accurately identifying or tracking individuals. This concept has been taken a step further with AI Camouflage, where designs created by AI are integrated into textiles.



  • Adversarial Fashion | Disrupting Surveillance 🧵



One innovative approach to protecting privacy in the age of pervasive surveillance is Adversarial Fashion. This refers to clothing designed to disrupt automated surveillance systems, such as facial recognition technology. By confusing the algorithms that detect and categorize human features, these garments help protect individual privacy.



  • Manifesto Collection | AI Camouflage 🦒


Cap_able brings advanced adversarial images from digital renderings to actual textiles using state-of-the-art knitting machines. Manifesto Collection was developed as a master's thesis in Design for the Fashion System by Rachele Didero, under the supervision of Professor Giovanni Maria Conti. Despite the complexity of the objective, the project aimed to create a knitted pattern due to the inherent properties of knitwear: increased wearability, comfort, practicality, and three-dimensionality.



  • Testing with YOLO 👀



The main goal is to prevent the faces of people wearing these clothes from being detected by surveillance software. This way, the software can't calculate their biometric data or create a facial signature. 


Tested with YOLO


Among the algorithms used for testing these garments is YOLO (You Only Look Once), one of the most widely used commercial object recognition software. YOLO is renowned for its speed and real-time capabilities, making it ideal for testing the effectiveness of adversarial fashion. The garments were tested with various versions of YOLO to ensure they successfully disrupted the algorithm's ability to recognize human features.  The system recognizes dogs, zebras, giraffes, or small-knitted people inside the fabric. ⭐️


  • Fashion's Role in Human Rights 🪐


As technology advances, our approaches to privacy protection must also evolve. Adversarial fashion and AI camouflage are innovative solutions that integrate privacy measures into our everyday lives. 


The wearer can choose which side of the garment to wear, allowing them to decide whether to activate the protective features against facial recognition technologies. This dual functionality is more than just convenient; it’s a deliberate design choice meant to encourage public engagement and discussion about privacy.


Tested with YOLO


Looking ahead, it is crucial to continue exploring the intersections of fashion, technology, and privacy. In a world where our faces are increasingly scanned, recognized, and recorded, safeguarding our most intrinsic data has never been more vital.



Pubblicato il  Aggiornato il