Self-hosted. Point a mic array (or several, PPS/GPS-synced) at the world and get a live map of what's making noise, where it is, and whether you should care.
curl -LsSf https://minimappr.com/install.sh | sh
One binary. A browser tab. Everything below runs locally on your hardware.
Full-bleed map, filtered live websocket feed, entity symbology, and uncertainty ellipses on every localized track.
Run one node standalone, or scale to a PPS/GPS-synced array with cross-node SRP-PHAT fusion and multi-server federation.
Identify bird species by call and place them on the map — not just a classification, a located, tracked entity.
Transcribe speech in range and fire alerts on matched keywords, with the triggering snippet attached as evidence.
Coyote detection ships as a default rule out-of-the-box; drones, gunshots, and anything else are supported via classifier labels plus your own rules.
Offline studio render pipeline exports spatial recordings to IAMF/MP4 for post-hoc listening and review.
ONVIF slew-to-track: point a PTZ camera at a localized sound event automatically.
RSSI multilateration to place Bluetooth devices on the same operating picture as audio tracks.
Per-node temperature/humidity sensing feeds live speed-of-sound correction for tighter localization.
Installs uv, then uv tool install's MinimapPR. Cross-platform, no manual Python setup.
curl -LsSf https://minimappr.com/install.sh | sh
irm https://minimappr.com/install.ps1 | iex
http://127.0.0.1:8080. To install without the BirdNET/YAMNet/ONVIF
extras, pass --base to install.sh. Start it any time with
minimappr; uninstall with uv tool uninstall minimappr.
MinimapPR's reference node design is a 4-mic tetrahedral array (PPS-synced for multi-node fusion). A full node build/setup tutorial is on the way — in the meantime, see the firmware README and node_config.h for node settings.