Generating your own captions .json

[ProgramItUp]

Greetings,

I'd like to create captions for my own images and save them in a .json similar to: http://msvocds.blob.core.windows.net/annotations-1-0-3/captions_train-val2014.zip with: images and annotations fields in the .json.

As a quick test with "python -m SimpleHTTPServer " was not successful, no data was written after submitting "http://localhost/coco_image_captioning.html

I looks like the configuration would be in mturk_form_submission.js but the specifics are a bit tricky.
Is anything else besides a webserver required to be running on the backend?

Suggestions welcomed.

Thanks.

[moormoon]

Hello,

Did you get a solution for creating coco segmentation annotations? I have the same question.

Thanks,