VlogMask: Tap to protect your face. Mosaic everyone else.

Hey Product Hunt 👋 I'm Kou, an indie iOS developer based in Japan. VlogMask started with a simple frustration: I love vlogging in Tokyo, but every time I filmed in a busy area like Shibuya, I'd spend hours afterward manually blurring strangers' faces in editing software. That can't be the answer in 2026. So I built VlogMask. It uses on-device face detection (YuNet CoreML + Vision framework) to find every face in the frame in real time. You tap the face(s) you want to protect — your own, your crew, your kids — and everyone else gets a pixel mosaic applied to their face automatically. The mosaic is burned into the recorded video, so what you see on screen is exactly what gets saved. A few principles I built around: 🔒 Privacy-first — Everything runs on-device. No video, no audio, no face data ever leaves your iPhone. Nothing to a server, nothing to me. 🎯 Profile + small face detection — Vision alone misses side profiles and distant faces (a real problem when bystanders are walking past). YuNet picks them up too. 📱 Native iOS — AVCaptureSession + Metal pipeline, real-time at 1080p/4K, front and rear camera, portrait and landscape, Bluetooth mic support. 💰 Honest pricing — Free works for solo creators (1 protected face, 5-min recordings). Pro is a one-time ¥1,000 / $7 unlock for unlimited recording and multi-face protection. No subscription. I'd love your feedback — especially from creators who shoot in public. What works, what doesn't, what's missing. Happy to answer any questions about the build (Swift, CoreML, on-device privacy approach, etc). — Kou kouworks