@user-c828f5 I'm also interested in the tons of paper for classifying eye tracking event data 😃

@user-29e10a haha sure, although I can't say I've read them all! Do PM me your email and I could forward a few papers?

I have been reading some papers from Nystrom and collaborators and they say that no single algorithm beats specialized human inspection in eye movements event detection. However they do not consider reinforcement learning and stuff like that. Anyone knows of a paper on AI and eye movements event detection?

@user-41f1bf These are two I have found recently... it's quite a new topic I guess

• Zemblys, R., Niehorster, D. C., Komogortsev, O., & Holmqvist, K. (2017). Using machine learning to detect events in eye-tracking data. Behavior Research Methods, 21(1), 1–22. http://doi.org/10.3758/s13428-017-0860-3
• Hoppe, S., & Bulling, A. (2016, September 8). End-to-End Eye Movement Detection Using Convolutional Neural Networks.

Zemblys et al. said they would be release the trained model openly, but their github account only contains the control-code

@matthias#6811 @user-41f1bf Truth is, human labellers themselves do not agree on events most of the time - which makes the problem even harder.

Rudra8, I would rather say "normal human labellers" do not agree most of the time. But specialized ones can achieve above chance agreements.

Hello all, I am currently in need of pupil recordings for my research. They have to be in situation where the observer is moving and gazing at something not necessarily stationary, but slower than himself. Examples woukd be jogging, driving/racing, riding a bike, trekking/climbing, etc.

Does anyone have any recordings that fit these criteria and can share them with me? You can send it to [email removed] or give me download links.

Thank you so much in advance

Sorry, I do not have any recordings that fits these criteria. I am working with stationary stuff.

@user-41f1bf thanks. But do you know if there is any kind of central repository for pupil recordings? I found some examples scattered through the website, but no link to a samples page. If I had access to a lot of recordings, I could sieve through them

Hummm, I don't think so. But it sounds like a good idea. A central repository with sample recordings would be well received. I think @mpk @wrp @papr would be happy if people uploaded their recordings to somewhere for free.

here are some http://dm.dgu.edu/link.html

maybe you find something here too: http://vintage.winklerbros.net/Publications/qomex2013eye.pdf

they have a lot of stuff, please have a look: http://ti.uni-tuebingen.de/Perception.perception-engineering.0.html

Wow, thanks a lot, @user-29e10a !! I'll check them out

Hi, any pointers to tools that allow the analysis of eye-gaze sequences, e.g., AOI1, AOI2...?

@user-11dbde yes there is the OGAMA project. They have Leveinsten and some other stuff in there

Pupil have some basic inside outside event detection for surfaces also

But Pupil have no analytics, just raw data.

hi thank you

i know ogama very well

however what I am looking for is something more advanced in terms of detecting and visualizing most frequent sequences of events

search of subsequences...etc...

Well, you are looking for a programming language

lol

yes...that is what i have done

and some machine learning

but i was wondering if there is any new tool on the block...after all this time...

What kind of tool?