Recognized by offers a mobile app that applies locally‑processed adversarial perturbations to photos, making them appear unchanged to human viewers while rendering them unrecognizable to facial‑recognition algorithms. The technology subtly shifts facial geometry by a few pixels, protecting users’ biometric data without uploading images to any server. It works in real time, allowing seamless integration into everyday photo‑sharing workflows.
Funding
Funding not disclosed
Founders
Product
Problem
Photos shared online can be automatically scraped and processed by facial‑recognition systems, creating biometric profiles without the subject’s consent. Existing privacy protections are weak, allowing companies and governments to build and sell databases that track individuals across services.
Solution
Recognized By offers a mobile app that applies local adversarial perturbations to images, altering facial geometry by only 1–5 pixels—imperceptible to humans but enough to confuse facial‑recognition algorithms. The processing runs entirely on the user’s device, so original photos never leave the phone, and the altered image looks identical to the original to human viewers. Users can edit and share photos instantly, ensuring their biometric identity remains private while maintaining the visual quality required for social media or professional use.
Target Audience
Primary users are privacy‑concerned individuals who share photos on social media, messaging apps, or professional platforms, as well as organizations that need to protect employee biometric data when publishing visual content.
Features
- Real‑time on‑device processing that applies AI‑driven adversarial perturbations without uploading data
- Human‑identical output: altered photos appear unchanged to human eyes, preserving visual quality
- AI‑invisible transformation that disrupts key facial‑recognition features such as inter‑ocular distance and jawline descriptors
- Simple mobile interface that integrates into existing photo‑sharing workflows
- No reliance on external servers, providing privacy‑by‑design and eliminating data transmission risks