Beamster
Nominee of the Talent Award 2025
Description
Beamster AR (Augmented Reality) Laser Tag is a unique kind of computer game – one that brings players together in the same physical space.
Inspired by the classic game of tag (“Fangenspielen”), Beamster reimagines this timeless concept using smartphones, computer vision, and AI. Instead of tagging someone by touch, players “catch” opponents by taking a photo of them with their smartphone.
If a player successfully captures another in an image, advanced Computer Vision and Person Re-Identification models confirm the tag by matching the photo to reference images taken before the session – what we call the “scanning process”. Once a player is tagged, one health point is deducted from their score. Each player starts with 10 health points, and each successful hit counts as one point.
Beamster uses XMPP, a robust messaging framework common in multiplayer games, to handle real-time communication between devices and the game server. The current MVP (Minimum Viable Product) version runs on both iPhone and Android and supports one game mode: Last Man Standing. More game types are planned, such as a mode where the “tag” is passed to the last player hit – similar to the traditional version of tag.
Games can be played one against everyone else or in teams (e.g., players can add a team tag like “Alex [Team A]” to their nickname). The game duration is configurable (e.g., 5 minutes), and a live leaderboard is displayed at the end.
We call the game “Beamster AR Laser Tag” because laser tagging is the closest real-world analogy. Laser beams are visualized on the screen, and laser sound effects accompany each shot. When a player is hit, segmentation is used to determine the impact area, and the hit is visualized with an augmented reality effect – a laser burst that appears to explode on the player’s body.
At this year’s JKU Games 2025 event, we hosted a “Beamster Laser Tag tournament”, which was a great success. You can watch highlights in the videos included with this submission.
.
Educational Institution
JKU (Johannes Kepler University) Linz
Program
Master Program Artificial Intelligence
Studio Name
Beamster
Product Page
Genre
Multiplayer Ego Shooting in RL
Platform(s)
Mobile