AayushKrChaudhary / RITnet

This is a winning model of OpenEDS Semantic Segmentation Challenge
MIT License
73 stars 28 forks source link

Hi, we are interested in your work, can you give me a OpenEDS dataset download link? #7

Closed huaji98 closed 1 year ago

huaji98 commented 3 years ago

We found no public download link for this dataset.

Lukeskywalker147 commented 3 years ago

https://cs.rit.edu/~cgaplab/RIT-Eyes/

huaji98 commented 2 years ago

https://cs.rit.edu/~cgaplab/RIT-Eyes/

Thanks, But this format does not match the code?

mixxen commented 2 years ago

Looks like the data in the link is simulated eye data. Would it be possible to get access to the real Open Eye Dataset?

AayushKrChaudhary commented 2 years ago

We do not have the right to share the OpenEDS-2019. The dataset is from https://research.facebook.com/openeds-challenge/ Regarding RIT-Eyes the format of the file will image loading script has to be changed to load those images.

perphyguo commented 1 year ago

Thanks for your dataset! After opening the .p file, I am confused about the details. Is there some place for me to learn he exact meaning? What I am doing is to locate pupil with normal camera, not in VR. So what is the data in 'pupil' mean? Seems it is not x, y coordinate in the picture. Thanks!

Capture

robbiedood commented 1 year ago

Hi @perphyguo , I was wondering how would you open these .p files. I have tried multiple apps but still can't make it. Hope to hear your expertise. Thx.

Thanks for your dataset! After opening the .p file, I am confused about the details. Is there some place for me to learn he exact meaning? What I am doing is to locate pupil with normal camera, not in VR. So what is the data in 'pupil' mean? Seems it is not x, y coordinate in the picture. Thanks!

Capture

perphyguo commented 1 year ago

Hi @perphyguo , I was wondering how would you open these .p files. I have tried multiple apps but still can't make it. Hope to hear your expertise. Thx.

Thanks for your dataset! After opening the .p file, I am confused about the details. Is there some place for me to learn he exact meaning? What I am doing is to locate pupil with normal camera, not in VR. So what is the data in 'pupil' mean? Seems it is not x, y coordinate in the picture. Thanks! Capture

I use a library to open .p file in python

robbiedood commented 1 year ago

@perphyguo thanks for your advice, we finally figured out the .p file is pickle file instead of matlab P-file :) !

perphyguo commented 1 year ago

@perphyguo thanks for your advice, we finally figured out the .p file is pickle file instead of matlab P-file :) !

So would you tell me the meaning of the labels if you figure it out?

Thanks!

lukelu312 commented 1 year ago

To my knowledge, labels are about segmented area for corresponding eye regions, or classification, (e.g. if eyeglasses exist), or camera geometric info, e.g. position. Which label you'd like to know the most ?
@perphyguo

AayushKrChaudhary commented 1 year ago

The pickle file you are reading is not the correct file for this model. The model only takes image as input and the label must be image with segmentation labels stored in numpy fashion. Using appropriate files may help to solve the issue

The OpenEDS dataset must be downloaded from the source owners [Me as the authors of this paper do not have legal right to share the dataset. Required format: Image as input Labels same size as input image but corresponding to the segmentation labels

shivaz560 commented 1 year ago

Hello. Could you give me an example label folder for one of your synthetic datasets ?