Eye Contact in Video Conferencing (2018)
itotd.com
itotd.com
On video chat sometimes I’ll say something like “mmhmm” or “yes” in agreement, and accidentally throw the speaking party off because my comment was delivered after a normal pause in speech due to latency.
This interruption of normal conversation flow is more problematic to me than eye contact. We both know were speaking into machines after all.
As a side note, I believe eye contact and 3D projected body language tracking are the limiting technology for virtual offices to leapfrog physical offices.
It’s my expectation that when those two technologies are cheap ($600 all-in-one device) it will massively disrupt both commercial real estate and any residential markets that are commute dominated.
I believe particularly in the Bay Area this will be the source of the next housing market collapse. Particularly since Facebook is heavily invested in telepresence now, and has a clear path to a device like I described within ~5 years.
Google and Apple, in all likelihood, will put out a similar device in that timeframe as well.
I have seen a few headsets with downward facing cameras but not many. There are hard problems around both technologies so they are mostly research prototypes at this point I think, not consumer alphas.
As I said I believe Facebook, Google, and Apple all have this tech on the far side of their roadmaps. So I don’t see all three failing.
But even if they do, the Leap Motion roadmap points there, and... while Leap is obviously mismanaged, new parent company is UltraHaptics. Hopefully they can get Leap into a headset....
Regardless, the Leap founder (Ed: David Holz) knows exactly what’s necessary for telepresence in VR, and I don’t think any amount of mismanagement will stop him from executing it EVENTUALLY.
But he’ll take 10 years, and Facebook might do it in 3. Apple could roll it out next year for all we know. They’ve prototyped all the tech already in ARKit.
David if you’re in this thread and you want to build a leap headset email me. We can just disassemble Oculus Go’s and add a Leap and... well you wrote the 6DOF code already right? Solder a coprocessor on the board that can do it without any battery drain. Maybe we have to design one, IDK. Sacrifice a kitten. Hire a stats PhD and make it “stochastic”. Put a neural net processor between the cameras and the computer. Who knows.
Then ship it with a Nintendo 64 emulator ONLY, that’s the entire OS, no games you have to sideload them. And mark it up to $999. Fund the company on preorders.
Needs IPD adjust, off ear audio
You would still benefit from having a point location in the meeting room. That’s one of the basic improvements VR gives you over videoconferencing. Maybe two floating eyes (for gaze direction and eye contact) plus some canned emoji-style gestures?
My point is you, as a person for whom current gen VR is enough—no eye or body tracking needed—are atypical. Most workers need body language and eye contact to communicate effectively.
Some of us, specifically blind people, have lived our lives without ever being able to make eye contact, yet we still have meaningful communication, both with each other and with sighted people. I, for one, don't ever think about my inability to make eye contact, except when a sighted person brings it up (edit: and to be clear, that only happens in a discussion like this one, not when trying to communicate with me personally).
that's just how things work. the same way folks without smell don't have any particular need for it, while for others smell creates intense emotional connection. just because something is not important to you doesn't mean it is not important to others, and vice versa.
is the article taking a fairly narrow view of communication? sure. it's video conferencing - the whole point is that people are looking at each other... but not quite. it's an uncanny valley! (a metaphor which probably works in multiple if not all senses)
i wonder how/if parallax eye contact impacts people using sign language?!
The reason I'm asking is that I do agree with the article, but I'd phrase it slightly different: The technology right now makes video chats ~different~ and awkward due to the lack of eye contact, IFF you're used to eye contact in face to face conversations in the first place, because video chats are supposed to allow f2f over remote distances.
I currently work in an on-site office environment, and only a small number of us are blind or low-vision. When I'm initiating a real-time conversation with someone, I'll often do it by IM. We may do it face-to-face if the other person is right there, but to me that's just like an audio call with 3D positioning, perfect fidelity, and no latency. To be fair, that is indeed still better than current VoIP, but not enough so that I would insist on being in the same building as my coworkers. Indeed, with an online audio group chat, I always know when someone joins or leaves the group; that's not always the case in person.
At my previous job, by contrast, we were an all-distributed team made up almost entirely of blind or low-vision people. As you would expect, we never used video. But we did use audio a lot, both one-on-one and group chats. (Our favorite voice chat system ended up being TeamTalk [1].) I never met most of my coworkers in person, and I never felt that this was a problem. It only just now occurred to me that the few sighted people on our team might have been uncomfortable with the way the rest of us worked. As for the rest of us, as far as I know, none of us ever wanted to be together in an office building. At least not for the sake of communication; having a shared pool of test hardware would have been nice from time to time.
[1]: http://bearware.dk/
When people capable for eye contact talk to each other and are not having eye contact then it creates an impression of not being there, or not valuing the conversation enough, or not wanting (or not being capable) to tell the truth.
Current video conferencing system creates an environment where participants do not necessarily feel comfortably. It is not equal to the real life conversation.
You can mitigate this by looking to the camera when you are talking and then to the screen when you are listening. This creates an impression that you are looking into the eyes of the listeners but then you can not truthfully see their reactions.
Never assume that just because you have a particular personal preference that most people share it.
This would be a crazy difficult (and fun) domain to work in. Humans are very well tuned to perceiving facial details - the margin for error would have to be tiny. I'm still uncertain if images processed this way could escape the uncanny valley, but this appears to be a much more technically feasible approach for consumer use than transparent displays, multi-cam setups, or camera arrays embedded in displays.
I'm as insecure about my appearance as anyone else- what's the deal? I keep my camera on until it's obvious that nobody else is going to show their face. Then I have to be the awkward one to turn off.
I do this. My camera is always broken.
I’m fine with audio, don’t like cheapening face-to-face meetings with video, and don’t see the reason to get ready for a remote meeting to which a video of my face adds zero value. It’s a preference which costs nothing, so I exercise it.
(If it’s a negotiation, why would I give my facial expressions away for free?)
I personally get annoyed when people video mute by default, and when I was a manager it was something I would bring up in 1:1s. Lurking on meetings is fine, but if you have a real role in the meeting and you aren't multitasking a meal or something, I feel strongly you should have your video on and be engaged, and most of the time when people are video muting it's because they are hiding their distraction and lack of attention, or else it's reached a critical mass where people are muting just to fit in.
Now that I'm an IC again generally the best I can do to encourage people to unmute is just name them: "hey Joe, are you there?" even though I know full well they are connected. It's admittedly a bit passive aggressive, but it's also a way of gently telling the person they aren't being present, and it's unlikely I'm the only person who notices.
I also mute myself for the wrong reasons, so I'm not trying to make this an indictment of individuals. But it's so easy for people to drift away from their work, especially when you are remote, and we need to look out for each other and not just accept dysfunctional behavior.
However, I can also see that my expectations are not shared by you, based on your strong response. My gut tells me this is related to different communication norms with "digital natives." I think your assumptions of malice and inability to learn on the part of the manager are misplaced.
I'll allow that some of it may be generational; many of my younger colleagues have never called in to a conference call.
But I challenge your claim that I'm assuming bad faith. At no point did I question the motives of the OP, only their actions.
Edit: I should clarify, because my original comment could reasonably be read as assuming details about the internal motivations of the OP. Allow me to rephrase: if a mangers' default response to an individual not having a camera on is to dress down the person for not "being present", that manager is making a toxic workplace environment.
Your preference seems... clear enough to me? given that you said you would quit that day. What am I misunderstanding?
I believe your response of leaving rather than attempting to reconcile the difference in expectations is an indication of assumption of bad faith. As in, you seem to believe the manager would not be responsive to a discussion.
Growing up pre-internet blessed me with the ability to communicate effectively without body language, thanks to the telephone. I find that translates perfectly well to video-less conference calls.
> most of the time when people are video muting it's because they are hiding their distraction and lack of attention
The OP assumes that those who have their camera off are wasting time, because they can't see them to prove that they're not.
A meeting is specifically about communication. A participant at the meeting that is inattentive is either wasting others' time, or should probably politely step out, or at least should not have been asked to attend. Being present is a priority for a good meeting, hence the pejorative "butts in seats" comment doesn't make sense to me.
You shouldn't be projecting your personal opinion on others. Even if people don't pay attention maybe you should consider that it's because the conference is not that important to them. Happens to me all the time. If I paid full attention in every meeting and call I am supposed to attend I would get nothing done.
I would not enjoy working for someone who thought like that.
Because of this, unless I'm expecting to be the primary presenter in a meeting or unless it's a 1:1, I'll usually "mute" the video feed. If I don't start video-muted, I'll frequently do it mid-call as people complain about lag, choppy audio, or inability to make out the primary presenter's screen. There's no reason to put the extra technical tax on the call just so people can see what I look like when I'm staring at a screen.
Please consider this as an alternative motivation, rather than assuming that anyone who isn't making themselves visible is slacking off.
If is a team call or a small group trying to solve a problem, the chances to have the camera on are higher.
In essence, the camera gives you a more intimate feeling, and depending on the situation people are more or less inclined to it.
Shouldn't the verb be "blinding" in this case?
I'm not sure why this is. I think it's not helped that subconsciously, video feels one way. A participant can be looking right at me but I can't know that, because I don't know where my face is relative to their camera. So, I assume that all participants are watching me all the time, even if that's an irrational assumption, and even though 99% of them won't care if I scratch my nose or something.
You could collect training data by setting up webcam through teleprompter, regular webcam and Tobii on a machine. Just record normal usage and then learn a model using eye-tracker data and regular webcam data as input and data captured through teleprompter as target.
With Tobii the system would also know who you are looking at on the screen. With custom conferencing software, you could deliver the software modified eye-contact stream only to that person.
There's no magic in how people look when they look at you. All the magic of feeling when someone is looking at you happens in the recipient's brain.