Udacity-Computer-Vision-Nanodegree-Program icon indicating copy to clipboard operation
Udacity-Computer-Vision-Nanodegree-Program copied to clipboard

Download and prepare the MS-COCO data set using Tensorflow

Open vidhi-mody opened this issue 5 years ago • 4 comments

Resources: MS COCO dataset

Note: This issue will be implemented in a separate branch. The PRs will not be merged in master.

vidhi-mody avatar Jun 27 '20 20:06 vidhi-mody

I'd like to take this one!

ChaoticBlack avatar Jul 04 '20 04:07 ChaoticBlack

@ChaoticBlack sure!

vidhi-mody avatar Jul 04 '20 06:07 vidhi-mody

@ChaoticBlack do you need help with this?

vidhi-mody avatar Jul 10 '20 13:07 vidhi-mody

I am facing an issue that i'm unable to figure out. I want to store the path of all the individual images in a list. If I use os.listdir, the cell takes too long to run and I dont get the expected results. If I use full_coco_image_path = PATH + 'COCO_train2014_' + '%012d.jpg' % (image_id) (this uses the JSON files to load the data) then certain images are not found. I re downloaded the dataset yet i'm facing the same issue.

ChaoticBlack avatar Jul 12 '20 05:07 ChaoticBlack