@everyone - eye tracking add-ons are now available for Microsoft HoloLens and The Epson BT-300: Checkout our post: https://pupil-labs.com/blog/2017-03/hololens-and-bt300-eye-tracking-add-ons/

Hi there! I have a couple of questions about the hmd-eyes calibration for the HTC Vive. I've been trying to get the calibration to work but I've ran into a couple of issue. Firstly, what actually is the procedure? I see the white and back concentric circle in the centre, then it moves to the top, top left, left and so on. Secondly, when the targets rotate around the field of view, I can barely see them, in fact I can only the targets on horizontal and vertical planes if I view them in my peripheral, and the targets on the diagonal I know are present due to a slight white blurring in the extremes of my vision in the direction of the target. And help would be most appreciated 😃

Hello everyone. I have problem with the calibration in the hmd-eyes unity integration when using the new 0.9 capture software. The eyetracking seems fine on the capture side (better than in v0.8), but the calibration process is broken when triggered from the unity side (i only get seemingly incorrect tracking for one eye). i get the following error in the capture side when i finish the calibration process.

world - [INFO] calibration_routines.hmd_calibration: Stopping Calibration world - [WARNING] calibration_routines.hmd_calibration: No matched ref<->pupil data collected for id0 world - [WARNING] calibration_routines.hmd_calibration: No matched ref<->pupil data collected for id1 world - [ERROR] calibration_routines.hmd_calibration: Calibration failed for both eyes. No data found

Thanks for the report. We will investigate!

Hi guys

As you know we're interested in buying the Hololens eyetracking add-on

We want to use the Hololens world camera alongside

I've done some reading, and as far as I understand we'll be able to use the Hololens' camera if we can provide images using one of the available backends (https://github.com/pupil-labs/pupil/tree/master/pupil_src/shared_modules/video_capture)

The NDSI backend looks most suitable, so I had a look in the repository and found this example which provides images using this backend: https://github.com/pupil-labs/pyndsi/blob/master/examples/uvc-ndsi-bridge-host.py

My idea is now to re-implement something very similar to that example, but publishes images from the Hololens rather than from a UVC device

Obviously I will need to use C# instead of Python as the Hololens only supports C#

Do you guys think it's a good way to go forward?

I'm nearly done in re-implementing the example

Hey @user-dfa26d

The example is really more like a test host and is not very cleanly implemented. It might be out of date as well. :/ But I think that generally it is a great use case for ndsi :)

Okay .. glad to hear ndsi sounds like the right way to go forward

Do you think you could update the example according to the latest changes?

I can have a look at it in the next days. The real reference should be the protocol definition though.

Sure, sure .. but going from an example is lots easier than going from the real reference (at least to me 😉 )

Yeah I understand that. I will ping you here when I made sure that the example is up to date

Great, thanks a lot!

BTW, where would the best place be to share the C# project as soon as it's finished? Guess this might be interesting for others as well

Either as a pull request to the hmd-eyes repository or in an own repository

Interesting. What do you mean by "edge detection"? Some sort of pigmentation?

Or agglutination?

I am sorry. These posts were not supposed to be posted here. I am sorry.

I had a feeling about it. :-D

@user-612f00 we fixed your bug. Please try version v0.9.2

@user-dfa26d I updated the host script to Pyhton 3 (commit: https://github.com/papr/pyndsi/commit/64c1545b1f18234ae5b318ad5b2e633587f284e0) but did not have to change anything protocol related. This means that you should be good with translating the old one

thanks a lot im gonna try as soon as i can tomorrow

Okay its working again. Thanks so much!

Great to hear that @user-612f00

Great thanks a lot @papr!

Hey people

Hope you are all well today

Hello @user-d82dd4 - greetings 👋

So i actually have a question about Pupil Unity intergration

as when i try to use it in the Calibration unity scene i only get input from one eye

has anyone had a similar experience? I will say however that the software on this computer is not exactly recommended, Windows7 and Unity 5.3

@mpk - this was recently resolved was it not?

I think I also need to take a closer look at the SDK though, as I dont think its being used atm because Unity hasnt been giving me this fabled "+" symbol mentioned in the readme on github

@user-d82dd4 - are you running Pupil Capture or Pupil Service on Window7?

Ive been running Capture

Interesting - and it is working on Windows 7?

yeah, quite well actually

it was mentioned in the driver download that it was made for Windows 10, but i figured why not and just went for it

I did not think/expect for it to work on Win7. We bundled on Win 10 and didn't think backwards compatability would be so friendly

in Capture it actually works just fine, both eyes are being detected and its all quite accurate with the little FOV camera on top

heh, well what do ya know, in capture it works well

that's good to know.

Regarding the other issue - I believe that this issue was raised previously and was resolved in the current master branch of Pupil Capture. But I want to double check on that with @mpk

aah alright, so should i just download that version real quick then i guess

And see what happens

@user-d82dd4 - please wait a few minutes for @mpk to chime in here

Sure, no problem!

have any suggestions come to mind about the Unity Pupil thing where it only detects one eye?

please use the latest version v0.9.2!

also please run calibration. If should show both eyes!