I recently rediscovered `leap motion` now called `...
# linking-together
d
I recently rediscovered
leap motion
now called
UltraLeap
. They have a version that attaches to a VR display that looks really nice Stereo IR 170. But what I found even more amazing are their haptics devices PS. Should we have a HCI channel? I'm going to post other links in the thread
In their newsletter they shared this 3D display from Sony

https://www.youtube.com/watch?v=KrLMnQM0_Psβ–Ύ

k
Add a new channel only if you want to have fewer people seeing what you post πŸ™‚ I'd say keep it here where it'll have more discussion. I've been watching David Attenborough describe a watering hole where otherwise solitary animals come to socialize: https://en.wikipedia.org/wiki/Mbeli_Bai
🀣 1
πŸ™ƒ 3
amiga tick 3
r
I have one of the original leap motion devices. It is terrible, constantly looses the track, but I always thought it had a lot of potential. I was quite sad when they shifted away from consumer facing products. It's ironic because the hand tracking built into the Oculus Quest is very similar, and much more robust.
πŸ‘ 2
w
I'll second that Quest tracking is good enough for the main menu: when I pop on the headset to download updates or whatever before beginning a serious session.
d
@Ray Imber did you find any interesting usages for the leap motion? @wtaysom is the tracking in Quest good for something else beyond the main menu?
w
Not that I've tried. The API was first experimental and still new.
r
I found the experience very hit and miss. the API was...limited... as I said before, the tracking algorithm lost fingers often enough that the experience was frustrating to use. And you couldn't get at the underlying data to tweak anything. You only got access to the final gesture detection or a limited ik hand model. I did a few internal toys, like using my hands to drive a presentation at the job I was in at the time. I also tried to use it for surface tracking instead of full 3D tracking, but unfortunately, I wasn't very successful there either. The unreliability made it a nice parlor trick, but impractical. This was about 4 years ago, bear in mind. I haven't given up on the concept of hand tracking. The algorithms continue to improve, and I think we will see a lot more cool stuff in this space in the future:

https://youtu.be/MbZ0ld1ShFoβ–Ύ

d
Thanks for the video @Ray Imber, it’s really cool. We just need that the technology gets from 10min/frame to real-time πŸ˜†
amiga tick 1