CrowdScope: Crowd Counter

How many people are really there? CrowdScope gives you a defensible answer in seconds. MEASURED ACCURACY · ON-DEVICE AI • Powered by a state-of-the-art crowd-counting AI model, running in real time on your iPhone's Neural Engine • About 12 % average error on dense crowds, measured on ShanghaiTech, the reference academic benchmark • Works offline: no server, no upload, no waiting • An honest range with its margin of error, never a made-up exact number Point your iPhone at a crowd: an AI model built for dense crowds estimates the number of people live, entirely on your iPhone. You get an honest range (e.g. 560–720 people) with its expected margin of error, not a made-up exact figure. BUILT FOR PROFESSIONALS Journalists, law enforcement, event organizers, venue and safety teams: when a number will be quoted, it has to stand up. KEY FEATURES • Live count with a density heat map • Sweep mode: turn on the spot to count a crowd wider than the camera, without counting anyone twice • Procession mode (CrowdScope Pro): iPhone on a tripod, count the people who cross a line during a march, both directions, for hours, with setup assistant and saved report • LiDAR filter: rejects false crowds such as patterned rugs or wallpaper • Automatic model choice for dense or sparse crowds • Screen / picture mode to count a crowd shown on TV or in a photo • Saved readings with date, place, GPS position and viewing direction • PDF report: range, map with viewing sector, anonymized images, reliability notes, reading number • Faces anonymized by default in reports PRIVATE BY DESIGN All processing happens on your iPhone. No image is ever uploaded. No account required. TRY IT FREE 3 free scans to test it on a real crowd. CrowdScope Pro unlocks unlimited scans and PDF reports (monthly or yearly subscription). Subscriptions renew automatically unless cancelled at least 24 hours before the end of the period. Manage them in your Apple Account settings. Terms of Use (EULA): https://www.apple.com/legal/internet-services/itunes/dev/stdeula/ Privacy Policy: https://franckultima.github.io/crowdscope/privacy.html