Prepare half resolution tfrecord datasets and use SegmentationTFRecords on Erins dataset where image suffixes are ".tif" and not ".tiff".
How did you implement your changes
Added to new arguments to SegmentationTFRecords: resize and img_suffix. Resizing is done with cv2.resize in SegmentationTFRecords.prepare_example and SegmentationTFRecords.get_inst_binary_masks. Suffixes are just added wherever images are loaded from disk.
What is the purpose of this PR?
Prepare half resolution tfrecord datasets and use
SegmentationTFRecordson Erins dataset where image suffixes are ".tif" and not ".tiff".How did you implement your changes
Added to new arguments to
SegmentationTFRecords:resizeandimg_suffix. Resizing is done withcv2.resizeinSegmentationTFRecords.prepare_exampleandSegmentationTFRecords.get_inst_binary_masks. Suffixes are just added wherever images are loaded from disk.Remaining issues None