Hi ! I would like some information. Is the HTC VIVE Binocular Add-on compatible with the XR Elite Headset ?

thanks ! If I purchase one or multiple add-ons from you, I need to calculate the customs fees since I live in Canada (Quebec). To simulate the amount of taxes, I need to provide the type of material. Among the choices are: "Cameras", "video consoles", "video games", "printers", "software", "computers", "batteries", "tablets and e-readers", "phones", "televisions/projectors/monitors". In which category do VR headset add-ons fall? Thank you.

OK; Thanks!

Dear PupilLabs Team,

I hope this message finds you well. Our names are Noah and Layne, and we are working on a research project on enhancing road safety through eye-tracking technology.

We are interested in knowing if PupilLabs is interested in exploring the possibility of incorporating their eye-tracking solutions into our research. Could we connect with a representative to discuss potential collaboration opportunities and pricing details?

Thank you for your time, and we look forward to your response.

Layne and Noah

Dear PupilLabs Team, My name is George and I have very stupid yet simple question about Pupil Core Open Source Code. How do I run it? I have read some instructions about the virtual environment required, yet I am unable to run it from the source. Are there other sufficient instructions that I am unable to find? Thank you

Hi, I wonder if the Pupil Core uses Infrared Oculography or Video-Oculography（VOG）？

Hi, I have a Pupil Invisible and I'm trying to export my recordings. I don't have the gaze.csv in my exported files. Is this something I have to generate manually?

Hi, I'm trying to stream the video and data from a raspberry Pi to another computer. The whole system was developed and worked perfectly before, but it sat for months without running, it failed to stream videos. I could still start pupil capture from the other computer, but just no videos were received. Please see the image attached.

Any tips would be greatly apperiacted!

Hi developers. Thank you for developing this amazing peice of hardware. We are integrating this eye tracker in our driving simulator. In a very plain language, this eye gaze data is used to manipulate the user interface on the windshield of the self-driving car. We have defined multiple area of interest (i.e., surfaces for gaze mapper). The data retrieved has a lag of couple of second. As a result, there is a lag in the user interface. We have a high performance PC: 64 gb ram, RTX 3080, Intel I9 12 core, and 10 GB graphics card. Is this a performance issue as a result of not up to mark hardware we have, or is it because we have define multiple surfaces (apx 8 surface) leading to a lag in data being sent to the subsribers? Let me know if you have any questions (ping me). I would really appreciate your response. Edit: we have 14 QR code trackers in our environment Edit: I am running the pupil eye tracker from source code and we have made a little modification to the data type of data being sent (only for a specif surface) to the subsribers. We are aware that this could lead to performance decrement, but we think it is not the major contributing factor. Edit: I have overlapping surfaces too

Hi, my name is Ken. When i transfered recordings via USB, how i can get "Gaze Overlay" movie (mp4 file) as showed in Neon Companion applications.

The driving simulator software we are using btw, is very expensive in terms of computation.

Have you examined the CPU load and RAM usage while running both Pupil Core and the driving sim? If you're maxing out the system, you may consider moving Pupil Core to a separate machine and streaming the data over the network

Thank you for your response. I will get back to you when I go back to my research lab and confirm if that is the case.

Hello

I have a question about the pupil labs eye camera. We want to connect it to an MCU, and we were wondering what cable it uses for data transfer?

Hi . I would like to use a USB extension cable to extend the length of the Pupil core data cable. Is this okay? If possible, are there any recommended USB extension cables? thanks

Hello Pupil team, recently we ran into several cases where Pupil Capture keeps detecting tiny "pupils" on participants eye lashes and eye lids. (See pics) The problem is that fake pupils can be detected even when the real pupil is visible. We initially thought that it could be women's mascara confusing the algorithm, but we've observed this in cases where the participant did not apply make-up. Note: the experiment environment is a low-light environment (as we want to measure pupil dilations) and the computer screen is the brightest source of light. I tried to change ROI but it does not always work. Can you give us some advice on how to overcome this? Thanks.

We are considering a Raspberry Pi MCU.

For the sake of clarity, the Raspberry Pi is not an MCU. It's a single-board computer (SBC). It consists of a CPU, RAM, and storage - like a very small PC. As such, these types of devices run typical mainstream operating systems (like Linux, Android, Windows, etc). These also typically have similar I/O capabilities as desktop PCs, like USB, HDMI, ethernet, etc plus sometimes general purpose I/O (GPIO) and bus interfaces like SPI, I2C, etc

Microcontrollers (MCU) are a separate class of devices. They typically have significantly less memory and do not run full-blown operating systems - instead only running a solitary program. For connectivity, these typically are limited to GPIO and bus interfaces like those mentioned previously - so usually no USB, no HDMI, no ethernet, etc.

The Raspberry Pi is a single-board computer (SBC), and the Raspberry Pi Pico is a microcontroller (MCU)

More specifically, what kind of wire connection does the eye camera use to be connected to an MCU or SBC?

Core eye cameras + MCU

Hi @user-5346ee! To expand on my colleagues answer, mascara can influence the tear film, exacerbating reflections that appear on the pupil images. This can negatively impact on pupil detection.

That said, there does seem to be decent contrast between the iris and pupil in the image you've posted.

Would you be able to share an example recording with us such that we can provide more concrete feedback? You can do so via [email removed] Please share the entire recording directory.

I am still a little confused about this, I do not have a gaze_positions.csv, only a gaze.pldata which I see that there is the "gaze_point_3d" and "eye_center_3d", are these the same as the gaze_point_3d_x and etc described in https://docs.pupil-labs.com/core/software/pupil-player/#raw-data-exporter?

and this gaze_point_3d_z - z position of the 3d gaze point is the optical axis?

also how do I open the world.intrinsics file? I assume this is where the camera intrinsics are located

Hello Pupil team, after analyzing some pupil invisible recordings and exporting them into iMotions, it has come to our attention that the gaze maps are off centered for many of the recordings. There is nothing in eye motion that can be done to fix this, but I was wondering if there was a way to correct the off centered data in the original recordings in pupil cloud? Is there a specific feature that would let me pull off the gaze map from the videos or alter the positioning of the gaze map on the video? Thanks!

why this is popping up ? my MacOS version is Sonoma 14.0

Hi, @user-870276 - for technical reasons, Pupil Capture on MacOS 12 and newer has to be launched with elevated privileges. More information and instructions are here: https://docs.pupil-labs.com/core/software/pupil-capture/#macos-12-monterey-and-newer

Hi @user-6072e0 ! One way is to use the surface output image ( the detected one, to perform the OCR) , effectively maintaining the same coordinates. Ain’t that an option for you?

setting with movement involved

We are analyzing eye-tracking data that was recorded using your AR/VR add-on for a VR simulation users are playing through an HTC Vive Cosmos. We want to be able to generate heat maps using the surface tracker plugin in Pupil Player but we are having difficulty enabling that feature. Specifically, it says we cannot add surfaces. Are we doing something wrong? Can we get more insight into how to generate heat maps using Pupil Player from a pre-recorded dataset?

@user-c2d375 I'm with Julianna responding to your query above. We did not have Apriltag markers in the VR scene. How do we set these tags up? Is it something that can be done only before the recording?

Hi, I have a technical issue with the device. The eye tracker seems not connect to the apps and doesn't show the display for the glasses and no data were collected.

Hi, @user-29f76a - if you're on a Mac, you'll need to run Pupil Capture with elevated privileges. See https://docs.pupil-labs.com/core/software/pupil-capture/#macos-12-monterey-and-newer for more information. Otherwise, please tell us a little more about your setup (what operating system, what version of our software, have you used it successfully in the past, etc)

i use the phone that comes with eye-tracking glasses (Neon Module and Frame). But the thing is, the glasses cannot connect to the phone @user-cdcab0

like this

Hi, while collecting the data with the phone provided, it shows up an recording error and I think it seems the file limit almost reach the limit of the phone provided. Could I remove the files manually in the phone? Would files already uploaded in the pupil cloud still stay after I manually remove that file in the hone? Thank you for your help in advance!

Is there any documentation on how to add events in the pupil cloud? I can't seem to see how you move the start and end markers. Also, how do you add several event markers to the timeline? thank you for any advice.

Hi Team! I have multiple gaze mappers (calibrations) in one video. So when I’m choosing the most accurate gaze mapper for my experiment among with them, do you recommend me to go for the one with the least accuracy result in post-hoc validation or the gaze mapper (calibration) that is temporally the most approximating to my experiment section in the video (so that less change could potentially happen between them, e.g., glasses slippage)? Can you help me please? Thank you very much in advance!!

Does Pupil Core require that the source of IR illumination is from the LED on the glasses? In other words, would the eye tracking still work if an external IR light was used instead? I figure that traditional eye tracking algorithms that detect the angle between the pupil and the glint from the IR light wouldn't work, but I don't know enough about the Pupil Core algorithm to know if the same problem would occur

Additionally, what wavelength of IR does the illuminator produce?

Hi Lency,

Happy to help. We could organise a meeting this week and I'll run through my steps?

Hello everyone! We are having problems when calibrating... Now it cannot find the "stop" mark at a distance of two meters, has anyone had the same error? Thank you

@user-d407c1How should I disassemble the Pupilcore, it looks like the cables are fixed inside the skeleton.

Hi @user-15e3bc ! Why would you like to disassemble Pupil Core? We can offer you a de-cased version if you would like to prototype

Sorry, I didn't think about integrating with other AR devices at the time of purchase. Is it easy to provide the connection for the de-cased version then? Or can the cable be supplied separately?

Please contact sales@pupil-labs.com to get a quote for a VR de-cased version.

Hello everyone,

I have a question regarding testing the pupil tracking system on a subject. After performing the Screen Marker calibration, I observed the following values: angular accuracy - 2.8 and precision - 0.61. However, even after this calibration, the tracker was unable to properly recognize the pupil. I'm wondering if this issue might be related to the pupil dilation, as one of the subject's eyes is more dilated than the other.

Hi, I have a question about the eye tracker method. The eye-tracking method is Infrared OcculoGraphy-based or video-based. Did the Pupil Core use a Infrared light source？

is it based on the principal that, if a fixed light source is directed at the eye, the amount of light reflected back to a fixed detector will vary with the eyes position?

Hey Neil! Thank you so much! Will try the setting and let you know, and are there any constrains when using the pupil core for low vision or people with defective eye? And which device you suggest me to use in this senario as I have (Pupil Core & Invisible).

🤔. I don't see a gaze circle in that screenshot. Just to clarify, your recording does contain gaze data, right? If not, that would explain the empty surface tracking export.

Do the cameras each work individually? If they do, and if your computer has more than one USB controller, you might have success putting them on different busses.

Hey folks. I believe I remember there is a way to force a monocular gaze map for individuals with strabismus, or who just have a bad track with one eye. Can someone please share that script?

What is the intensity of the IR illuminators on Pupil Core (in either W/sr or W/cm^2 on the eye)?

Hello, I am working with a team on an academic project designing a DIY eye tracking system. We have come across the 120 Hz individual camera for the Pupil Core and are curious if our team can get ahold of a data sheet for this camera to see if it would be beneficial for our application as opposed to the de-cased webcams used in Pupil Lab’s DIY instructions.

Additionally, can you use a single 120 Hz camera or is it designed to be used in pairs?

We already did experiments on eye and hand tracking on normal people doing their daily activities. Now we wanted to do it on people with low vision or defective eyes to observe their gaze and fixations while doing daily activities which doesn’t involve drastic movements. In this process we are coming across subjects with bigger pupils or subjects having defective eyes and the Pupil core isn’t able to obtain the gaze and fixation data in these situations. So what would be a better solution for this?

Hi @user-edef2b! Check out this message for reference: https://discord.com/channels/285728493612957698/285728493612957698/1163435437445087322

I meant Irregular eyes or eye with cataract. And what do you suggest using for this application pupil core or pupil invisible?

Hey, regarding Core and Capture, I'm currently refactoring / remaking the LSL plugin to work with surfaces as well. Upon debugging, I noticed that the gaze event list isn't ordered by timestamp. Though I realize that by being timestamped the problem is solvable on the receiver's end, I'd like to send the LSL samples in their correct order. Is there any reason they aren't sorted?

Or does this happen when there is generally low confidence?

Also, is there a way to access a plugin from another plugin? (not refering to the notifications)

Hello, can anyone recommend antiseptic wipes/ spray to clean the Core headset that is available in the UK? Thanks!

Hi @user-74fd00 ! Have you seen our recommendations for disinfecting Pupil Core? https://docs.pupil-labs.com/core/hardware/#disinfecting-pupil-core

Hi, here is a link to recording that i made with pupil core. I wanted to know what would be the reason for poor confidence values and improper gaze detection https://drive.google.com/file/d/1CwJsn0FUMkMThFFA_K3dFHA3P7B-E-6m/view?usp=drive_link

@user-79be36 Hi! Your colleagues has let us know that we did not respond to one of your questions here in discord, and that you needed assistance.

Would you mind letting us know what is your question either here or at info@pupil-labs.com ?

Hello everyone! I just installed pupil capture on linux but it seems the device could not connect. Any suggestions?

Hello, is there a wiring diagram for the Pupil Core right and left eye sensors? I want to wire the sensors to my own Raspberry Pi SBC. Additionally, what is the power draw for each sensor (how much amperage does each sensor use)?

Regarding the timestamps and the events, do they generally come unorderered? Or can it be attributed to low confidence? I typically see timestamps in recent_events that are later than timestamps that come in the next list of recent_events.

Hi Neil. Thank you. What is the power draw for each sensor?

Hello everyone ! I'm using pupilcore with LSL for an experimentation. I collect the data via LSL and via pupil capture, using the LSLRelay plugin. I have the gaze positions recorded via LSL in my xdf file, but also the .csv generated by pupilPlayer after processing the data acquired via pupil capture. I noticed that the timestamp associated with the LSL data goes backward sometimes (fig 1 shows the timestamp diffs with negative values ) while it does not with the processed pupilplayer file (fig 2, also timestamp diffs). I therefore wonder 1/ Why do I have these negative values in the LSL file and 2/ how does pupilplayer adjust the timestamp to obtain only positive diffs ? Many thanks for your answers !

Also, what kind of USB is needed to connect the sensors to the Pi (2.0, 3.0, Type-A, Type-B, etc.)?

Hi

This is Manuel from Universidad Politécnica de Valencia (Spain)

I have pupil w120 e200b

and I can not finish the calibration step

as soon as yon can, could do you help me, pls?

in the 2nd image, you can see the calibration stop at 4th movement

Hi @user-a81678 ! Could you please navigate to the folder pupil_capture_settings, copy the capture.log after the crash and share it with us at [email removed]

ok. I will go

I can´t find this folder or file in Pupil v3.5.1 folder

@user-d407c1

not finish/detect the marker

Ok! then we might not need the log file, could it be simply that the scene camera can't see the marker? perhaps you need to rotate it slightly down?

@miguel(pupli Labs)

Miguel, it works now like you have told me!!

thank you!!

Hi This is Karan here from Universitat de Barcelona. We are using Pupil core for our neuromarketing experiment.

I had certain doubts regarding the data analysis.

We tried a lot but cannot understand how to proceed with the data analysis.

Also, can we make heat maps with the pupil labs free software?

Please let me know!

Thanks 😊

Hello! I wanted to learn where I can find the serial number for my pupil core device. Can I see it on my hardware or through the pupil capture software?