def extend_with_flipped_entries(roidb, dataset):
"""Flip each entry in the given roidb and return a new roidb that is the
concatenation of the original roidb and the flipped entries.
"Flipping" an entry means that that image and associated metadata (e.g.,
ground truth boxes and object proposals) are horizontally flipped.
"""
flipped_roidb = []
for entry in roidb:
width = entry['width']
boxes = entry['boxes'].copy()
oldx1 = boxes[:, 0].copy()
oldx2 = boxes[:, 2].copy()
boxes[:, 0] = width - oldx2 - 1
boxes[:, 2] = width - oldx1 - 1
assert (boxes[:, 2] >= boxes[:, 0]).all()
flipped_entry = {}
dont_copy = ('boxes', 'segms', 'gt_keypoints', 'flipped')
for k, v in entry.items():
if k not in dont_copy:
flipped_entry[k] = v
flipped_entry['boxes'] = boxes
flipped_entry['segms'] = segm_utils.flip_segms(
entry['segms'], entry['height'], entry['width']
)
if dataset.keypoints is not None:
flipped_entry['gt_keypoints'] = keypoint_utils.flip_keypoints(
dataset.keypoints, dataset.keypoint_flip_map,
entry['gt_keypoints'], entry['width']
)
flipped_entry['flipped'] = True
flipped_roidb.append(flipped_entry)
roidb.extend(flipped_roidb)
def extend_with_flipped_entries(roidb, dataset):
"""Flip each entry in the given roidb and return a new roidb that is the
concatenation of the original roidb and the flipped entries.
"Flipping" an entry means that that image and associated metadata (e.g.,
ground truth boxes and object proposals) are horizontally flipped.
"""
flipped_roidb = []
for entry in roidb:
width = entry['width']
boxes = entry['boxes'].copy()
oldx1 = boxes[:, 0].copy()
oldx2 = boxes[:, 2].copy()
boxes[:, 0] = width - oldx2 - 1
boxes[:, 2] = width - oldx1 - 1
assert (boxes[:, 2] >= boxes[:, 0]).all()
flipped_entry = {}
dont_copy = ('boxes', 'segms', 'gt_keypoints', 'flipped')
for k, v in entry.items():
if k not in dont_copy:
flipped_entry[k] = v
flipped_entry['boxes'] = boxes
flipped_entry['segms'] = segm_utils.flip_segms(entry['segms'],
entry['height'],
entry['width'])
if dataset.keypoints is not None:
flipped_entry['gt_keypoints'] = keypoint_utils.flip_keypoints(
dataset.keypoints, dataset.keypoint_flip_map,
entry['gt_keypoints'], entry['width'])
flipped_entry['flipped'] = True
flipped_roidb.append(flipped_entry)
roidb.extend(flipped_roidb)
def flipped_roidb_for_training(roidb):
"""Load and concatenate roidbs for one or more datasets, along with optional
object proposals. The roidb entries are then prepared for use in training,
which involves caching certain types of metadata for each roidb entry.
"""
flipped_roidb = []
for entry in roidb:
width = entry['width']
boxes = entry['boxes'].copy()
oldx1 = boxes[:, 0].copy()
oldx2 = boxes[:, 2].copy()
boxes[:, 0] = width - oldx2 - 1
boxes[:, 2] = width - oldx1 - 1
assert (boxes[:, 2] >= boxes[:, 0]).all()
flipped_entry = {}
dont_copy = ('boxes', 'segms', 'gt_keypoints', 'flipped')
for k, v in entry.items():
if k not in dont_copy:
flipped_entry[k] = v
flipped_entry['boxes'] = boxes
flipped_entry['segms'] = segm_utils.flip_segms(
entry['segms'], entry['height'], entry['width']
)
flipped_entry['flipped'] = True
flipped_roidb.append(flipped_entry)
roidb.extend(flipped_roidb)
logger.info('Computing bounding-box regression targets...')
roidb = filter_for_training(roidb)
add_bbox_regression_targets(roidb)
logger.info('done')
_compute_and_log_stats(roidb)
return roidb
