HTC today announced it’s releasing both facial and eye-tracking units for its Vive Focus 3 standalone VR headset.

The company says its new trackers will offer “more natural and immersive VR experiences, new options for enterprises to measure user responses, and even more accessible interaction methods such as gaze control.”

The units are priced at $100 for the Facial Tracker and $250 for the Eye Tracker, however that’s pretty much a drop in the bucket in comparison to Vive Focus 3’s business-oriented price of $1,300.

The new Vive Focus 3 Facial Tracker captures expressions through what HTC says are “38 blend shapes across the lips, jaw, cheeks, chin, teeth, and tongue to precisely capture true-to-life facial expressions and mouth movements on avatars.” The unit, which attaches to the headset’s USB-C port, is built around a mono tracking camera clocked at 60Hz.

Image courtesy HTC

Developers will be also able to implement both facial and eye-tracking in Vive’s Wave SDK and soon via OpenXR, with integration available in Unity, Unreal Engine, and Native.

SEE ALSO
Update: PSVR 2 to Include Tech from the Biggest Name in Eye Tracking

The Vive Focus 3 Eye Tracker is also a plug-and-play solution, fitting magnetically into the headset and conveniently allowing for the headset’s normal IPD adjustment range.

Like in many such aftermarket setups, the insert is studded with IR illuminators around the lens and fitted with dual cameras facing the user’s eyes, which are said to capture data for gaze origin and direction, pupil size and position, and eye openness.

Image courtesy HTC

HTC is touting some of the big benefits of eye-tracking in general with the device, with the most obvious being the ability to increase social immersion. There’s a long list of benefits to eye-tracking though, including foveated rendering, hands-free UI manipulation, and not to mention the mountain of data you can infer from the user’s gaze direction, duration of gaze, pupil response, etc.

Although you may think HTC is playing catchup by offering the aftermarket bits now to match Meta’s upcoming Project Cambria (likely Meta Quest Pro) spec for spec, this isn’t actually the first time HTC has offered the tech in its headsets.

In 2021 the company released a consumer version of its Vive Facial Tracker for Vive Pro and Vive Pro Eye, the latter of which came stock with eye-tracking courtesy of Tobii. This will however be the company’s first in-house aftermarket units, with third-parties such as Droolon, 7invensun, and Pupil Labs having filled that particular gap with bespoke eye-tracking devices.

This article may contain affiliate links. If you click an affiliate link and buy a product we may receive a small commission which helps support the publication. See here for more information.


  • Andrew Jakobs

    Has anyone actually ever bought/used the facetracker? Are there games/apps that use it?
    HTC has a lot if hefty discounts lately on all their headsets, or a free wireless module. I thought it seems like they are clearing their Vive Pro and Cosmos stock, so maybe there is a new headset from HTC around the corner, the Flow is still not completely targeted for regular VR

    • Christian Schildwaechter

      Ever bought: apparently yes. I went to the HTC developer forums looking for information about their eye tracking, and found developers asking the same question. It seems a lot of VR/gaming studios bought the face tracking add-on to use them for avatar eye and lip animation, and there are enterprise projects using data from eye and face tracking. Plus the usual VRChat enthusiasts that can use it via OSC. The tracking module was actually sold out for some time.

      Games/apps that use it: besides VRChat pretty much only professional in-house/custom use.

      • Arno van Wingerde

        But I cannot see it going much further than that – unless the Cambria makes this such a prevalent option that it actually pays for developers to actually do something with it! This has always hampered use of pretty much any add-on, like a gun where you put your controllers in and such. Good for Chatting and development, typically not worth it for most of us…

        • ViRGiN

          the point of integrating such things into a headset is to give early ‘devkit’ access for actual developers to pick on. this will obviously make it into more affordable quest series in the future. unlike literally every company out there, meta products are actually purchased and used. nobody cares when valve releases ‘finger tracking’ controllers, nobody cares when pimax launches eye tracking, nobody cares when htc releases trackers – but when meta does something, everyone picks on it. hand tracking, mixed reality – it all already got adopted.

          • silvaring

            So the point of this new technology is to track peoples eye and mouth movements in real time? Sounds pretty creepy eh

        • Christian Schildwaechter

          Cambria is part of Meta’s bet on social VR, which so far is still a (large) niche inside a small niche. Success of facial tracking will mostly depend on the success of social VR. The costs for facial tracking hardware shouldn’t be high, even with the wide optics they use integrating an IR camera plus LEDs should be less than USD 10. The XR2 can already handle seven camera inputs, so technically the Focus 3 can natively deal with four HMD, two eye and one face tracking camera, with each of them taking some to a lot of the processing power.

          We expect future HMDs to add eye tracking anyway, because it has multiple other uses for interaction and performance, but if facial tracking proves to even attract a small number of users, HMDs might start to include it just for that. And then other use cases will be found, it could for example assist with voice recognition, so if the sound is hard to distinguish, lip position could be used as a second source. You can also derive a lot of information from a persons face, a game detecting the user frowning and moving the jaw or biting lips could determine that s/he is currently stuck and offer a hint.

    • Guest

      These companies are just drooling over the mountains of data they can get from eyetracking and think people are stupid enough to pay for some Mythical benefits from it!

      • NL_VR

        What are those “Mythical benefits”.

        • Guest

          There’s a link in the article for those drinking-the-koolaid. It’s an 8 or 9 tailed fox!

          • NL_VR

            you need a foilhat or something?

        • Christian Schildwaechter

          For the advertisers: Eye tracking is sort of the wet dream of any advertiser. Combined with A/B testing where you display two alternative ads to a number of users and determine where the eye moves first.

          Usually they can only determine what the users finally clicked, not how they got there, and that alone can lead to very frightening results and is the cause for all the clickbait we see. Now imagine them knowing what you gaze was automatically drawn to, even if you resisted clicking it. We’d inevitable end up with personalized virtual world flashing in exactly the right shapes and colors to subconsciously draw your attention.

          Unless you are a professional poker player, your whole facial expression is pretty much a traitor that can get you into a lot of trouble. So far mostly with other humans, but in XR with eye and face tracking with data center algorithms too.

          For the users:
          – more performance through dynamic foveated rendering
          – easier interactions, we already saw examples of better physics in VR through eye tracking when throwing things, because the game can determine what you were actually aiming at
          – better interaction in social VR
          – move towards the similarly mythical metaverse, which might be a turn of for many VR players, but a lot of people have seen appeal in the cool Hollywood version in Ready Player One

          • NL_VR

            you consider it mythical benefits?
            it sounds to me he meant there be no benefits maybe i understand wrong.

          • Christian Schildwaechter

            The benefits for the users are so far unproven. We have been hoping for fast improvement from eye tracking with dynamic foveated rendering for years, with little evidence that it will actually have a large impact. Improvements for social physical interaction have been shown in experiments, but we will have to see if they are really useful beyond some niche cases. And even the dangers of user manipulation are currently only seen at a lab level.

            This may not be really mythical in the sense that these don’t exist, but none of them are proven to be actual impactful and therefore worth pursuing. We see this with a lot of technology. I’m a big fan of 3D printing, but none of the world changing predictions made about it turned out to be true, as the technology is still to primitive and labour intense and requires acquiring a lot of skills to be actually useful. So the revolution turned out to be more of a very useful niche technology only a very tiny group of people actually uses.

            And one can argue that VR itself is full of these mythological benefits, where in the early (current) phase from 2012-2016 many believed we’d soon get close to the Oasis, if only the technical problems would be solved. There are of course still a lot of these, but the tech improved a lot. Nonetheless even with pretty decent and affordable HMDs like the Quest 2 with lots of actually good software, thanks to heavy subsidization by Meta, it turns out most people just don’t care. Even among those that have actually tried it, and worse, a lot of people who actually bought one stop using it after a few months. It too turns out to be a niche technology only a very tiny group of people actually uses.

            This may of course all change, but we often jump onto tech for promises that turn out to be not as compelling in reality. Hence mythical benefits.

          • XRC

            Bill Gates, 1996 “We always overestimate the change that will occur in the next two years and underestimate the change that will occur in the next ten.”

          • NL_VR

            Verry true

          • Christian Schildwaechter

            I broadly agree, but this is also used as an excuse to stick to overly optimistic projections. Technically we are now beyond the ten year mark, with the current wave of consumer VR pretty much being kicked off by John Carmack showing Palmer Luckey’s VR prototype at E3 in June 2012, shortly after id software announced that they’d support HMDs in a future Doom 3 BFG update, which made pretty big waves. And I’d say most people who jumped the hype train back then actually expected more would have changed by 2022.

          • NL_VR

            Yes you may be right but i dont agree its not worth pursuing.
            We that was the mentality we wouldnt have anything.
            We will soon se if foveated rendering give performance boost as Sony have implemented it on PSVR2 just of that reason.

          • Andross

            “new options for enterprises to measure user responses”
            well, I always tought about it, they just confirmed the thing like it’s a normal thing.

            no fucking way. we need some laws, but looking at our politicians it’s more problably they will shout “witchcraft!” or ” gamble games!” and ban all the VR products at all.

  • Tiny Focus fan-base rejoice! Your terrible headset is getting an amazing feature… just slightly ahead of much, MUCH better headsets… a few months from now.

    If I was absurdly rich, I would get a Focus 3 just to use this for those few months. Still $1300 vs $300 Quest 2 (soon to be $400). That’s a hard sell, even if you are a multi-millionaire!