Loading video...

Video Failed to Load

Go Home

A developer figured out how to access Quest 3's passthrough cameras to run custom computer vision models. Details here:

44,945 views • 2 years ago •via X (Twitter)

7 Comments

Rael Hodgson's profile picture
Rael Hodgson2 years ago

Proud to work on this team. Always thinking outside the box 🎁

Nova 🐧✨'s profile picture
Nova 🐧✨2 years ago

oh my god video encode then decode and using a compressed video stream for CV? this thing sure is horribly inefficient (Facebook's fault)

Robi ᯅ's profile picture
Robi ᯅ2 years ago

nice, i'm using the same approach and can verify it works like that! :)

skibidiwap69's profile picture
skibidiwap692 years ago

@cixliv 🧐

Secondson Of Swan's profile picture
Secondson Of Swan2 years ago

@cixliv I mean its not meta giving access but it's something

Bozoid's profile picture
Bozoid2 years ago

amazing MR not only allows me to see real reality now but also tells dumbass me what I'm seeing!

Mahdi Jahangiri's profile picture
Mahdi Jahangiri2 years ago

I hope this becoms standard access for devs. MR "needs" camera info to be MR otherwise it's a glorified versoin of AR. (Eg, you can actually remove objetcs and modify them much better than whats now possible with AR games. rn quest apps cant access the camera.)

Related Videos