I’ve been interning at Bosch where I created a toolkit to enable image detection on MR Headsets without main camera access like the quest 3 to enable things preciously not possible, like MR indoor navigation, colocation to mobile AR via smartphones as well as a surface detection replacement for quick marker less allignment to for example tables (to have for example all shadows line up correctly). Mechanism takes 3-10 seconds before allignment is finished, has regular image detection accuracy and works both for handtracking as well as controllers with no additional hardware required (also no jailbreaking and no performance cost after setup). We originally planned to publish it as a open source Unity Toolkit with a short accompanying Paper but with the recent news of Meta opening up main camera access next year, rendering custom solutions unnecessary I wanted to get get feedback if you think we’ll get image detection next year (see the long delay on avatar legs) and if yes, you all as developers would still be interested in something like that for the time before its release? Would love to hear you guys thoughts on it!
