BlurPOV detects every face in a video, clusters appearances by identity, and lets you choose — face by face — who to redact and who to reveal. Live from the camera, or applied to clips from your library.
Point the camera at a scene and BlurPOV blurs every face it finds in real time. Tap a face to flip its state, or set a global default — redact everyone, reveal everyone, exceptions per subject.
Pick a clip from Photos or Files. BlurPOV scans every frame, clusters faces into people, and shows you a gallery. Tap who to blur, render, and save the redacted video back to your library.
Vision-based face detection runs on every frame, generating bounding boxes and feature embeddings.
Appearances are tracked within shots and clustered across the video, so each person becomes one tile in a review gallery.
Tap thumbnails to mark redact or reveal. Drag-record manual blurs over moments the auto-detector missed.
BlurPOV bakes your decisions into a new video file and saves it to your Photos library.
Face detection, clustering, and rendering all happen locally on your iPhone using on-device machine learning. BlurPOV does not upload your videos, does not store your faces, and does not run analytics on your usage. The clip stays yours.
Protect bystanders and sources before publishing footage from sensitive scenes.
Share clips of your kids without sharing every other kid in the frame.
Use real-world video material in teaching and analysis without exposing identities.
Redact incidental faces from evidence or training material before it changes hands.