[email removed] do you have any suggestions for bibliography in research with eyetracking and bodily movement? we are recording movement of yoga but we can't find any previouse paper that can explain some paradigm about this kind of recording and calibration method...
Hi all! I`m currently starting a postdoc project on cognitive load in speech perception. Planning on using pupil dilation as a measure of effort while listening to various speech signals in quiet and noise. Have any of you done something similar with this equipment? Seems our first challenge will be to log information on pupil dilation on a timeline that also mark the events in the stimuli. Peder Heggdal, Haukeland University Hospital, Norway.
Hi I'm writing my Bachelor thesis and I have used Pupil for a eye tracking experiment. I got the advice on this channel that I should be able to achive an accuracy of 1-2 degrees when doing calibration. I got that, but I wonder if there is any reference that I can use that supports the claim? I'm also wondering if there is any reference that discusses the confidence level of the tracking? Like at what threshold it is reasonable to disregard the data?
Hey, Florian here from Ilmenau University of Technology, Germany. I'm working on perception in immersive environments, especially how different types (CAVEs, HMDs, mixed realities) compare to each other. Eye movement behaviour is one aspect of this π If anyone wants to catch up, let's keep in touch! π
@user-92dca7 could you please provide some references for @user-d72566 if you have them readily available?
@wrp I found a very good forthcoming paper--"A novel approach to single camera, glint-free 3D eye model fitting including corneal refraction " -- was written by Kai Dierkes; Moritz Kassner; Andreas Bulling.π But I can't read the paperοΌcould give me a complete paperοΌ Thank you!
I am a master, I am writing my master thesis.
π
Hey @user-c68bba, thanks for your interest in our forthcoming paper. π At the moment the manuscript is still in the last stages of the editorial process and thus not yet available to the public. I will let you know, when that status has changed. All the best for your master thesis work!
Hey @user-92dca7,thanks for your response.π
Hi, I'm using regular web cameras for both eye and world. When reviewing my footage however, I noticed that the FOV of the world view is somewhat limited. I am now thinking of buying a camera with a higher FOV (I think the one I have is about 50) and maybe a fisheye lense. Would this work with pupil or would it cause problems with the gaze tracking?
@user-d72566 you can do this, but you will need to calibrate camera intrinsics for your new optics/camera
Aha. I just tried following the instructions found on https://docs.pupil-labs.com/#camera-intrinsics-estimation for my current camera but I'm unable to follow the instructions. Step 4 says to resize the show pattern window to full screen. Then step 5 says to aim it at the pattern. Then step 6 says I should have the world view in focus, then what is the purpose of step 4? Cause the world view will now "block" the show pattern window. Then step 6 continues to say that I should press C, but that just triggers the normal calibration process for markers.
Hmm I think I figured it out, seems like the hotkey has been changed to 'i' instead. My front camera has a FOV of about 50 so I tried calibrating with the "radial" option. When calibration was done I ended up with a bunch of blue rectangular shapes. If I press "show undistored image" the image looks a little like a fisheye which allows me to capture more of the scene! Is this applied in the background or should I toggle this button when I want the fisheye view?
Hi, I have a question concerning the surface sizes. How can I set proper sizes without using the graphics interace? When entering 0, 0 for size, it still has the same old size. How can I change it using numeric values instead of dragging the corners in the world view?
Thank you!
Hi Tim,
the size is to set the real world size (think mm or px). to set the shape and size of the surfaces as they appear in the screen you need to use the handles in edit mode.
you should set the real world size to something meaning full for heatmaps and exports to work properly.
Hi mpk, thank you very much for your fast answer. I harldy dare to ask: But what are the "handles in edit mode"? How do I get there? Thank you!
Hi. Click on the button edit surface in the top right corner of each surface and then move the corners.
:)That kind of answers my question above too. Thank you very much !
@user-a04957 sorry, I did not read your first q fully. To rephrase, changing the position of surface verts in the marker coord system requires doing it via the handles.
Thanks again,
... It would be nice if you could answer me another question: Is it normal that the surfaces drift to a side when the viewing angle increases? What can I do to prevent this?
Thank you!