Photo Scrubber
An open-source experimental privacy tool developed by Simon Willison that performs face blurring and EXIF metadata removal entirely within the browser, requiring no server upload. It is built on Google MediaPipe and WebAssembly technologies.
Timeline (last 90 days)
Photo Scrubber is an experimental tool that can automatically detect and blur faces in photos while also stripping metadata, with all processing done locally in the user's browser without uploading photos to a server.
Photo Scrubber was created by developer Simon Willison, motivated by his desire not to share photos of protesters with identifiable faces of strangers on the internet.
Photo Scrubber is based on Google's MediaPipe C++ library, compiled to WebAssembly to run in the browser.
Photo Scrubber's face detection relies on the npm package @mediapipe/tasks-vision, combined with the BlazeFace face detection model to perform recognition.
Photo Scrubber is positioned by its author as an experimental tool, with room for improvement in detection accuracy, edge case handling, and other areas.
All Facts (5)
Photo Scrubber is an experimental tool that can automatically detect and blur faces in photos while also stripping metadata, with all processing done locally in the user's browser without uploading photos to a server.
50%UnverifiedPhoto Scrubber was created by developer Simon Willison, motivated by his desire not to share photos of protesters with identifiable faces of strangers on the internet.
50%UnverifiedPhoto Scrubber is based on Google's MediaPipe C++ library, compiled to WebAssembly to run in the browser.
50%UnverifiedPhoto Scrubber's face detection relies on the npm package @mediapipe/tasks-vision, combined with the BlazeFace face detection model to perform recognition.
50%UnverifiedPhoto Scrubber is positioned by its author as an experimental tool, with room for improvement in detection accuracy, edge case handling, and other areas.
50%