micahwalter edited this page Feb 27, 2012 · 11 revisions

In addition to the raw dataset, we are making available the corresponding media assets in the form of 350x350 pixel thumbnails. Media filenames are stored in the media.csv file which is a flat-file dump of our media table. This table includes the following fields:

  • id
  • filename

The id field is unique to each asset and is referenced via the media_id filed in the objects.csv table and elsewhere. You can use this relationship to match up media assets with their corresponding object records.

Each asset listed in the media.csv file is stored as a 350x350 pixel thumbnail on our media server and can be accessed via the following root domain.

Simply append the id ( or media_id from objects.csv ) to this URL and you should see the image in your browser as shown below:

NOTE: Please be aware that all URI's on Amazon S3/CloudFront are case sensitive. As such, we have chosen to use all lowercase characters for these asset locations.