Hi @nmt, unfortunately not, restarting both the browser (firefox) and laptop (on mac os) that I am using, and companion device did not resolve the issue. Just checked on another browser (chrome) without any luck!

How to generate a heat map

i unble to do

i am unable to do it

Hi! I'm trying to run visualization, but it continually shows error. How should I do to complete the analysis?

Hi! @nmt . I'm not sure if the message is providing more detail. In my case, I recorded about 5 minutes driving data, trying to run the visualization.

Hello, I'm trying to generate a heatmap and an error keeps popping up. I don't understand exactly what video or image that I need to insert into the Pupil Cloud that the heatmap can be generated.

I have a pupil labs glasses and i had not problem beforehand but now the recording won't get uploaded on the pupil could. (with invisible companion) Instead its stuck at 0% upload. i can pause and start again but nothing changes

cloud*

Ah, thank you. It worked ^^" @user-480f4c

Hi, Neil, thank you for the response. I attach here a print screen form the pupil cloud.

Hi, I'm using pupil invisible for a project. I'm having a problem, when i download a mapper maker from pupil cloud i noted that the file surface_positions has less rows than the gaze file and i can't understand why Is possible to link the data between these two files?

I also tried to download in pupil player format, elaborate with pupil player,but the fixation_on_surface file is empty

Thank you for the response

Good afternoon. I recorded 57 participants using Pupil Invisible (short ride on the tram sim). Then, in the Pupil Cloud environment, I defined 7 enrichments based on markers. Now, I need to check how many fixations each participant had in each area designated by the markers. The fixations file contains all the fixations of my participants. I have 7 enrichments, so I have 7 files. I believe that each file should contain the same number of fixations. The difference lies in the fact that in each file corresponding to a specific enrichment, there is a column with either true or false - depending on whether the fixation was present in that area or not. But numer of all fixations should be the same.

Hi @user-23899a ! Just to fully grasp it, you have a project with 57 recordings, and 7 enrichments inside with the same start and end events, but with different markers chosen, is that right? when you download the recording data in the fixations.csv you should have an id for the fixation matching the ones in the enrichments data.

If that's not the case, would you mind inviting us to your workspace so we can investigate it?

Could you add me to your workspace (||mgg@pupil-labs.com ||) so I can further assist you or send the files to data@pupil-labs.com ?

invitation sent

Hello, I have been having issues downloading the raw data export zip file from pupil cloud of my pupil invisible recordings. I have tried different browsers, wired and wireless connections, and different locations. The download starts and then stops after it gets to a certain point every time this has happened on two different videos as well.

@nmt It stops sliently no error

Recording downloads

Seems like reference image mapper doesn't work for files longer than 3 minutes. Any way around this? the visual helps!

Hi @user-91d7b2 👋 ! It seems there is a misconception, the limit of the reference image mapper is only for the scanning recording. Meaning, you can have one scanning recording of less than 3 min and then a longer one where your subjects wear the glasses for as long as required.

Hi, I'm using pupil invisible for a project. I download the pupil player format data in order to using pupil player (because the different frequency of the surface) and get the gaze on surface already scaled , but the number of rows in "gaze" (downloaded from pupil cloud) and the ones in "gaze position on surface" (from pupil player) sometimes aren't the same. So i'm trying to match the timestamps but i'm having a problem, the first "gaze timestamp" in "gaze position on surface" and the first "start timestamp" in "fixations" are different. How can i do?

Hi

We have started working with the PupilLabs Invisible for a few hours and sessions and a few days ago they stopped working. This includes not showing the fixation on the video, the application crashed (stops responding during recording) after a few seconds of seemingly recording - no video gets created. We have tried testing everything with the new pair that we just initiated today and that one too does not work at all. We are not sure why this suddenly happened and what could be the cause of the problem

We would like to ask for your assistance as soon as possible since we do not want our work to be delayed and this is an inherent part of our work

Thanks and best wishes Maayan

Sometimes we record and when we try to watch the video we get "No scene video", but in other times when we try to record the app just don't respond (we can't stop the recording and then we have to close the app).

Hello, can I copy or move some recordings to another Workspace in Pupil Cloud? I don't want all students to have an access to all other recordings. Kind regards and thank you.

I have a question about invisible companion. The app shows the two grey circles with the eye and the camera symbol. Sometimes the speed of the symbols going around the circle (its hard to explain sorry) differs (the recording time is accurate though). E.g. for example if i run my first program it takes one minute for them to pass one circle. Sometimes it takes three minutes (whe ni run the second program). Does that have any influence on the data or what is going?

However sometimes the second program shows the same pattern on the app as the first one. Or generally sometimes if i just press record on the app itself, it needs 3 minutes for one circle and other times one again

When i looked into the data, it looked normal to me but i don't know

@user-d407c1 Alright, if the speed doesnt matter, everything is fine^^. I just noticed sometimes it just moves differently (speed wise) but i never saw a gap or anything. And it also normally colors the grey cricles during recording So i guess it'll be okay. Thank you!

Hi @user-3c26e4 ! Sorry to hear, please upvote this feature if you would like to see it implemented in the future.

In the meantime, yes, you can create a visualisation and export it. To do so, add all recordings you want to create the visualisation to a project. Then, enter the project, and choose Analysis at the bottom left. Once on the Analysis page select "+ New visualisation" and choose "Video renderer". You will then be on a new page that allows you to modify the appearance of the gaze overlay, show fixations and even undistort the video.

Hi all.....I have 3 questions: 1) my one plus tells me to update the OS. Can I go ahead? will tha app work with the new android? 2) How can I move a video from a workspace to a new one? 3) don't know what happened. in the "recordings" sections of the app there isn't a recording I took. But i can find the video inside the Directory (see photo attached) thanks a lot

Hi I have recorded data using invisible and export them into folders and transfer them to PC. I want to know 1. how can I calculate some fixation state, saccade state, blink state, and eye moving metrics like velocity based on these data and Pupil Player? 2. I have deleted all the raw data on the One Plus app, so the cloud method can not be used, right?

I'm a bit confused... I'm using the invisible and reading in the movie and the bin data. But sometimes the video appears to be mirrored. Like I'm driving in England on my bike. Any ideas?

python open cv

I see... sorry for that

Hi @user-86d342 👋 ! You can use this CLI to compute blinks and fixations in the same way as Pupil Cloud does. With the difference that Pupil Cloud would reprocess the recordings and recompute gaze at 200Hz, and with this utility you will be limited to the sampling rate you got at recording times.

Hi, We are trying to make a recording with the invisible. But after 5 seconds a pop up shows asking to Failed to start recording, pleae disconnect Pl! Any ideas?

Hi, we recorded with Pupil Invisible at night. Unfortunately, the QR codes in the video are too dark and Pupil Player only recognizes them in some places. 1. the AOIs are only partially defined, so they jump around in Pupil Player when I play the video. (They deform depending on the view). Is there a way to minimize this? 2. even if the AOIs are only partially defined, I can download a table where they are all listed. Why can't they be displayed visibly in the video? 3. is there a way to define the frontal area for the AOIs in Pupil Cloud, even if the QR codes are too dark? Thank you very much.

Hello,

I am currently working on a project that involves using an 'invisible' device in a golf setting. This device is worn by players while they are on the golf course. I have several questions regarding its functionalities and capabilities:

Pupil Dilation Information: Does this version of the device provide any data or insights related to pupil dilation?

Time Tracking with Cloud Data: Is there a mechanism in place to determine timestamps or track time using the data we receive from the cloud?

Reference Image Compilation: Can we aggregate or compile zones of interest from 18 different videos to create a comprehensive reference image?

Neon Version Advantages: I am considering the Neon version of the device. Would this version offer additional information or enhanced features compared to the invisible device we are currently using?

Thank you for your assistance.

Hey guys, today i used the pupil labs glasses and something went wrong so i had to do two recordings when i wanted one. I wondered now if i can edit the first recording to the point where the data is usable so i had only download the raw data until that point. That would be more useful than searching for the timestamp and trying to delete all the data up to that timeframe from every dataframe. Is there a possibility to edit and cut a video until a point and afterwards merge it with the second recording? (i know i would need to change fixation ids and blink ids myself but i would like to edit the first video at least because the last seconds are not usable) Is there a possibility or do i have to do it coding wise or just try to delete in the csv itself by hand?

@user-d407c1 @user-480f4c really sorry to bother you guys...

do you have some possible solutions or explanations to the problems? can I expect this can be solved in one day or two?

@user-86d342 Scratch that I misread the logs. It is in the _process_world().

Could you try this branch ?

You can install it like this `pip install -e [email removed]

Then you can run it pl-rec-export -f your/path

We used 4 markers, but mostly only 2 to 3 are visible in the video.

Hi @user-cd7b85 , I removed my answer because I saw my colleague also replied to you.

As you can see our answers goes on the same line, if the markers are too dark to be detected, the marker mapper and surface tracker will struggle.

Detecting the markers is key to define the Surface/AOI, and usually at least 2 markers need to be detected to properly define them. If a marker is lost, then the surface definition will "jump".

A way to prevent this would be to place more markers across the cockpit/dashboard that can be recognised, making the surface definition more robust.

You can also try to better illuminate these markers, perhaps placing some kind of backlight?

If you really need this data, may I suggest a "hacky" way to try to save some data? Since you are using Pupil Player, you can try making a backup of the raw recording, then attempt to use FFMPEG to increase the brightness in the scene camera video files? That said, as my colleague mentioned the right marker seems to be partially occluded, so it may not work.

Thank you, I will try that. Is there a way to brighten up the video file in Pupil Cloud?

I have the problem that sometimes my invisible companion shows that i am recording (i am starting it through a software) but in the end i don't have a video or audio i can look at. One time it did work for the first program (but no audio or scene, also no error message in python/visual code) and the second different program was fine. And the second time neither first nor second program worked.

Also i have audio issues in general. It seems like the audio is sometimes really really quiet. We tried to take of the camera and put it back again and that worked for a while but now it's not working anymore. Sometimes the audio is load, sometimes its more quiet Do you know what could be the problem? Is it probably more hardware related?

@nmt Yes, i have a program by which is start recording and stop recording. And that normally works very well, except for these 2 times (within 24 times)

Hi @user-cd7b85! You can find it here.

The command you are looking for should be something like this: ffmpeg -i input.mp4 -vf "eq=brightness=value" output.mp4

where value should be from -1 to 1 , with 0 meaning no change

Hello, is there an example of an article about basketball players ?

We cannot operate the monitor system. Do you have a solution for this?

Hi, I'm currently trying to use "pl-dynamic-rim" through the CMD. During the process, the tool asks me for a "corners_screen" file, and I'm a bit uncertain about the format this file should follow. I've already interactively selected screen corners as instructed in the code but I want to run it through the CMD. Could you please provide guidance on the expected format of the "corners_screen" file?

Hi, I tried running the following code to check if the device connects to my laptop. But unfortunately the error appeared: "No device found".

device = discover_one_device(max_search_duration_seconds=15)
if device is None:
  print("No device found.")
  raise SystemExit(-1)

*** my laptop version is: HP Pavilion Laptop 14-dv0xxx Device name LAPTOP-TJJSK9DU Processor 11th Gen Intel(R) Core(TM) i7-1165G7 @ 2.80GHz 2.80 GHz Installed RAM 16.0 GB (15.8 GB usable) System type 64-bit operating system, x64-based processor Pen and touch No pen or touch input is available for this display

Windows: Edition Windows 11 Home Version 22H2

Python: Python 3.12

Dear Pupil Labs team. I am writing the methods part of a paper. Is there a document specifying, how fixations, dwell times, blinks and blink durations are detected and computed within Pupil Cloud/Pupils Invisible?