reductionista opened a new pull request #16: MADlib image loader: load dataset from disk URL: https://github.com/apache/madlib-site/pull/16 Adds to the madlib_image_loader module ability to load images directly from disk (instead of having to be in numpy arrays). New functionality can either be called as a module or run from the command line. (Loading from a numpy array still only supported when called as a module.) Dataset to be loaded is expected to be organized by image category/class. Each directory name labels a specific class, and only images from that class should be in that directory. An additional img_name column is added to the table when used in this way, to help with tracking which image is which later. Madlib Image Loader Demo notebook has been updated to include a minor interface change (original API endpoint renamed to load_dataset_from_np) and an example of loading from disk with newly added endpoint load_dataset_from_disk. Co-authored-by: Nikhil Kak <[email protected]>
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
