The Valve Index VR headset has been around for a few years now. It doesn’t come with eye or face tracking, but that didn’t stop inspired folks like [Physics-Dude] from adding DIY solutions in elegant and effective ways using a combination of hardware, open software, and 3D printable parts.
This project leverages the EyeTrackVR project (and optionally, Project Babble for mouth tracking) which both have great applications particularly in social VR spaces.
These are open-source, self-contained and modular solutions intended for a variety of hardware platforms. Of course, every millimeter and gram tends to count when it’s something that gets worn on one’s head, so [Physics-Dude] tailored a solution specifically for the Valve Index. His project makes great use of the platform’s hacker-friendly hardware design.
[Physics-Dude] also makes excellent use of a certain widely-available “gumstick” style USB hub as an important part of his build. Combined with with the front-mounted USB port on the Index, it results in an extremely compact and tightly integrated solution that looks great. While it can be risky to rely on a particular off-the-shelf item in a build, doing so absolutely has its place here.
The documentation is fantastic, including welcome guidance on cable routing and step-by-step instructions. If you’ve been interested in adding eye tracking to a project, be sure to give it a look. Already have eye tracking in a project of your own? Tell us all about it!
Still surprises me that the valve index is still one of the top tier PC VR headsets in 2024.
It’s a shame VR has kind of sputtered and died a bit, the Index is great kit to use but the small userbase of VR in general seems to create a chicken-and-egg problem with developers making high-quality games for it. I always meant to make an audio/microphone control board for the frunk, but never quite got around to it and I’ve not had the kit out for a while. Maybe I should give it a go again.
I don’t think it has sputtered or died myself, the explosive growth that happened when Valve+HTC first release the Vive and later when FB put out apparently good but very cheap headsets may have reduced a bit now, but it doesn’t seem to me like the interest has waned. More likely you have just caught up on the backlog of interesting to you VR applications and there isn’t enough new stuff to feel like its a vibrant and growing as there was when you still have heaps of things to try for the first time.
I have an original Vive, and also a (desktop) eye tracker from Tobii. Told myself I wouldn’t buy another VR headset until Eye Tracking is included and supported as standard in games. It’s such a natural fit, and I am surprised it’s been taking this long.
I’m not sure Eye tracking is really that important myself, it can certainly help reduce the computational load with foveated rendering (though on an OG vive there really won’t be much point, it doesn’t have the pixel count to really benefit), and it could be cool in some settings. But people in general interact with their hands, which VR is already good for so it seems like Eyetracking wouldn’t naturally improve the experience of using VR for most things. Unless your HMD has way more potential resolution than the computer can render, so it needs to know where to render sharp details.
combining gaze/eye tracking and hand interaction increases the bandwidth of the Human Computer Interface :)
foveated rendering does certainly allow higher framerate, which is important for comfort
there’s a new unaffordable headset that does both of the above, and future HMD will certainly improve upon both the function and price.
eye is a very high bandwidth interface, both in terms of perception (seeing) and expression (movement)
thanks for your input!
It certainly does increase the bandwidth, but other than the foveated rendering the eye tracking as an input seems like it wouldn’t be much use, at least to many. People are just wired to reach and interact with their hands, so unless you can’t and have to interact with the machine more via your eyes it doesn’t seem that important.
A value add still absolutely, but not worth putting off an upgrade from a Vive over if you can afford it. So for somebody who like myself still has a Vive, something so low resolution and not a super high refresh rate either that I’m pretty sure my Steamdeck could drive it just fine for most VR tasks… I keep meaning to try that out but pulling the desk out to liberate the HMD, while I expect the relatively low bandwidth and inevitably flakey USB-C connection will cause a few hiccup seems too much hassle unless it has to happen anyway.
Also there are a few comparatively affordable headsets that have at least one, and often both features (I’ve been looking on and off for an upgrade I can’t really afford for a while)… Perhaps not Vive, Index, or Occulus levels of affordable, but still not the fruit brand tax… Though in this case I think the price may actually be more justifiable than normal.
I think you need to meet more people with face tracking. There is a LOT of subtle non verbal communication that conveys so well with it (especially the time between thought and response where you can see the unconscious reaction)!
Meet people, at least to talk with them type interactions? You do that in meat space, or with a webcam if you must do it remotely surely, simpler and clearer…
VR is IMO much more for the stuff you can’t afford or practically do in the real world, like interacting with a giant full 3d hologram of your MRI scan, sculpt and product develop digitally with real depth perception to get the correct sense of scale, or simulator stuff. Though I guess that does mean my focus isn’t on that element where perhaps the it does add enough to be worth it.
Oh hey its me. Also shoutout to all the other people making/showing their versions of this build in the ETVR Dicsord’s DIY Showcase.