search

DiGyt / cateyes

Categorization for Eye Tracking - simplified
BSD 3-Clause "New" or "Revised" License
25 stars 10 forks source link

Doubts about nslr_hmm algorithm #9

Open giuseppecartella opened 1 year ago

giuseppecartella commented 1 year ago

Hi!

I have some questions about the usage of the nslr_hmm algorithm (classify_nslr_hmm() function).

I have x and y coordinates (in pixels) from eye tracking recording. The origin (0,0) is on the upper left corner of an image. Should I convert them to coordinates relative to the center and then use the cateyes.pixel_to_degree function?

Thank you Giuseppe

DiGyt commented 1 year ago

Hey Giuseppe 🙂

The NSLR-HMM uses angular data (=the angle of the eye relative to the center/fixation point) as input. If you have flat coordinates/pixels you can use cateyes.utils.coords_to_degree or cateyes.utils.pixel_to_degree to convert both gaze arrays to degrees.

Hope this helps 👍

DiGyt commented 1 year ago

Hey @beppefolder I just updated the code and the documentation for these functions. Hope they are more intuitive and easier to understand now. If they helped solving your problem, feel free to close the issue :)

giuseppecartella commented 1 year ago

Hi @DiGyt, At the end I had to start another work and I stopped the previous one. I 'll let you know in some days as soon I will recover the project. For the moment, thank you for the help and for enhancing the documentation.