def sample_mini_batch(self, anchor_box_list_gt, anchor_box_list,
class_labels):
with tf.variable_scope('avod_create_mb_mask'):
# Get IoU for every anchor
all_ious = box_list_ops.iou(anchor_box_list_gt, anchor_box_list)
max_ious = tf.reduce_max(all_ious, axis=0)
max_iou_indices = tf.argmax(all_ious, axis=0)
# Sample a pos/neg mini-batch from anchors with highest IoU match
mini_batch_panoptic_utils = self.dataset.panoptic_utils.mini_batch_panoptic_utils
mb_mask, mb_pos_mask = mini_batch_panoptic_utils.sample_avod_mini_batch(
max_ious)
# mb_pos_mask = tf.Print(mb_pos_mask, ['line 683(avod) : tf.shape(mb_pos_mask) =', tf.shape(mb_pos_mask)], summarize=100)
# mb_pos_mask = tf.Print(mb_pos_mask, ['line 684(avod) : mb_pos_mask =', mb_pos_mask], summarize=100)
# mb_mask = tf.Print(mb_mask, ['line 685(avod) : tf.shape(mb_mask) =', tf.shape(mb_mask)], summarize=100)
# mb_mask = tf.Print(mb_mask, ['line 686(avod) : mb_mask =', mb_mask], summarize=100)
# max_iou_indices = tf.Print(max_iou_indices, ['line 687(avod) : max_iou_indices =', max_iou_indices], summarize=100)
# class_labels: a tensor of shape [num_of_classes] indicating the
# class labels as indices. For instance indices=[0, 1, 2, 3]
# indicating 'background, car, pedestrian, cyclist' etc.
# For pedestrian training only, class_labels =[1]
mb_class_label_indices = mini_batch_panoptic_utils.mask_class_label_indices(
mb_pos_mask, mb_mask, max_iou_indices, class_labels)
# mb_class_label_indices : a tensor of boolean mask for class label
# indices. This gives the indices for the positive classes and
# masks negatives or background classes by zero's.
mb_gt_indices = tf.boolean_mask(max_iou_indices, mb_mask)
return mb_mask, mb_class_label_indices, mb_gt_indices
def test_iouworks_on_empty_inputs(self):
corners1 = tf.constant([[4.0, 3.0, 7.0, 5.0], [5.0, 6.0, 10.0, 7.0]])
corners2 = tf.constant([[3.0, 4.0, 6.0, 8.0], [14.0, 14.0, 15.0, 15.0],
[0.0, 0.0, 20.0, 20.0]])
boxes1 = box_list.BoxList(corners1)
boxes2 = box_list.BoxList(corners2)
boxes_empty = box_list.BoxList(tf.zeros((0, 4)))
iou_empty_1 = box_list_ops.iou(boxes1, boxes_empty)
iou_empty_2 = box_list_ops.iou(boxes_empty, boxes2)
iou_empty_3 = box_list_ops.iou(boxes_empty, boxes_empty)
with self.test_session() as sess:
iou_output_1, iou_output_2, iou_output_3 = sess.run(
[iou_empty_1, iou_empty_2, iou_empty_3])
self.assertAllEqual(iou_output_1.shape, (2, 0))
self.assertAllEqual(iou_output_2.shape, (0, 3))
self.assertAllEqual(iou_output_3.shape, (0, 0))
def test_iou(self):
corners1 = tf.constant([[4.0, 3.0, 7.0, 5.0], [5.0, 6.0, 10.0, 7.0]])
corners2 = tf.constant([[3.0, 4.0, 6.0, 8.0], [14.0, 14.0, 15.0, 15.0],
[0.0, 0.0, 20.0, 20.0]])
exp_output = [[2.0 / 16.0, 0, 6.0 / 400.0],
[1.0 / 16.0, 0.0, 5.0 / 400.0]]
boxes1 = box_list.BoxList(corners1)
boxes2 = box_list.BoxList(corners2)
iou = box_list_ops.iou(boxes1, boxes2)
with self.test_session() as sess:
iou_output = sess.run(iou)
self.assertAllClose(iou_output, exp_output)
def sample_mini_batch(self, anchor_box_list_gt, anchor_box_list,
class_labels):
with tf.variable_scope('pplp_create_mb_mask'):
# Get IoU for every anchor
all_ious = box_list_ops.iou(anchor_box_list_gt, anchor_box_list)
# all_ious = tf.Print(all_ious, ['-=-=-=-=-=-=-line 753(pplp) : all_ious =', all_ious], summarize=100)
# all_ious = tf.Print(all_ious, ['-=-=-=-=-=-=-line 754(pplp) : tf.shape(all_ious) =', tf.shape(all_ious)], summarize=100) # tf.shape(all_ious) = [4 265]
max_ious = tf.reduce_max(all_ious, axis=0)
# max_ious = tf.Print(max_ious, ['-=-=-=-=-=-=-line 756(pplp) : max_ious =', max_ious], summarize=100)
# max_ious =[0 0 0 0 0 0 0.225453332 0 0 0 0 0 0 0 0 0 0 0 0 0 0.417117327 0 0.404093683 0 0 0.00502740964 0 0.199674159 0 0.288573176 0 0 0.0159720983 0 0 0 0.169192255 0.00572701544 0 0 0 0 0 0 0.0461278595 0 0 0 0 0 0 0 0 0.128080159 0 0 0 0 0 0.615373969 0 0 0 0 0 0 0 0 0 0 0 0 0.309623659 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.0571997948 0 0 0 0 0 0 0 0 0 0 0 0.106632032...]
# max_ious = tf.Print(max_ious, ['-=-=-=-=-=-=-line 757(pplp) : tf.shape(max_ious) =', tf.shape(max_ious)], summarize=100) # tf.shape(max_ious) = [265]
max_iou_indices = tf.argmax(all_ious, axis=0)
# max_iou_indices = tf.Print(max_iou_indices, ['-=-=-=-=-=-=-line 759(pplp) : max_iou_indices =', max_iou_indices], summarize=100)
# max_iou_indices =[0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 3 0 2 0 0 2 0 3 0 0 0 0 2 0 0 0 2 2 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 3 0 0 0 0 0 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0...]
# max_iou_indices = tf.Print(max_iou_indices, ['-=-=-=-=-=-=-line 760(pplp) : tf.shape(max_iou_indices) =', tf.shape(max_iou_indices)], summarize=100) # tf.shape(max_iou_indices) = [265]
# Sample a pos/neg mini-batch from anchors with highest IoU match
mini_batch_panoptic_utils = self.dataset.panoptic_utils.mini_batch_panoptic_utils
mb_mask, mb_pos_mask = mini_batch_panoptic_utils.sample_avod_mini_batch(
max_ious)
# mb_pos_mask = tf.Print(mb_pos_mask, ['-=-=-=-=-=-=-line 768(pplp) : tf.shape(mb_pos_mask) =', tf.shape(mb_pos_mask)], summarize=100)
# mb_pos_mask = tf.Print(mb_pos_mask, ['-=-=-=-=-=-=-line 769(pplp) : mb_pos_mask =', mb_pos_mask], summarize=1000)
# mb_mask = tf.Print(mb_mask, ['-=-=-=-=-=-=-line 770(pplp) : tf.shape(mb_mask) =', tf.shape(mb_mask)], summarize=100)
# mb_mask = tf.Print(mb_mask, ['-=-=-=-=-=-=-line 771(pplp) : mb_mask =', mb_mask], summarize=1000)
# max_iou_indices = tf.Print(max_iou_indices, ['line 687(pplp) : max_iou_indices =', max_iou_indices], summarize=100)
# class_labels: a tensor of shape [num_of_classes] indicating the
# class labels as indices. For instance indices=[0, 1, 2, 3]
# indicating 'background, car, pedestrian, cyclist' etc.
# For pedestrian training only, class_labels =[1]
mb_class_label_indices = mini_batch_panoptic_utils.mask_class_label_indices(
mb_pos_mask, mb_mask, max_iou_indices, class_labels)
# mb_class_label_indices = tf.Print(mb_class_label_indices, ['-=-=-=-=-=-=-line 781(pplp) : tf.shape(mb_class_label_indices) =', tf.shape(mb_class_label_indices)], summarize=100) # tf.shape(mb_gt_indices) = [265]
# mb_class_label_indices = tf.Print(mb_class_label_indices, ['-=-=-=-=-=-=-line 782(pplp) : mb_class_label_indices =', mb_class_label_indices], summarize=1000) # tf.shape(mb_gt_indices) = [265]
# mb_class_label_indices : a tensor of boolean mask for class label
# indices. This gives the indices for the positive classes and
# masks negatives or background classes by zero's.
mb_gt_indices = tf.boolean_mask(max_iou_indices, mb_mask)
# mb_gt_indices = tf.Print(mb_gt_indices, ['-=-=-=-=-=-=-line 786(pplp) : tf.shape(mb_gt_indices) =', tf.shape(mb_gt_indices)], summarize=100) # tf.shape(mb_gt_indices) = [265]
# mb_gt_indices = tf.Print(mb_gt_indices, ['-=-=-=-=-=-=-line 787(pplp) : mb_gt_indices =', mb_gt_indices], summarize=1000) # tf.shape(mb_gt_indices) = [265]
return mb_mask, mb_class_label_indices, mb_gt_indices
def sample_mini_batch(self, anchor_box_list_gt, anchor_box_list,
class_labels):
with tf.variable_scope('avod_create_mb_mask'):
# Get IoU for every anchor
all_ious = box_list_ops.iou(anchor_box_list_gt, anchor_box_list)
max_ious = tf.reduce_max(all_ious, axis=0)
max_iou_indices = tf.argmax(all_ious, axis=0)
# Sample a pos/neg mini-batch from anchors with highest IoU match
mini_batch_utils = self.dataset.kitti_utils.mini_batch_utils
mb_mask, mb_pos_mask = mini_batch_utils.sample_avod_mini_batch(
max_ious)
mb_class_label_indices = mini_batch_utils.mask_class_label_indices(
mb_pos_mask, mb_mask, max_iou_indices, class_labels)
mb_gt_indices = tf.boolean_mask(max_iou_indices, mb_mask)
return mb_mask, mb_class_label_indices, mb_gt_indices
def test_iou_mask_ops(self):
# corners are in [y1, x1, y2, x2] format
corners_pred = tf.constant(
[[4.0, 3.0, 7.0, 5.0],
[14.0, 14.0, 16.0, 16.0],
[0.0, 0.0, 21.0, 19.0],
[3.0, 4.0, 5.0, 7.0]])
corners_gt = tf.constant(
[[4.0, 3.0, 7.0, 6.0],
[14.0, 14.0, 15.0, 15.0],
[0.0, 0.0, 20.0, 20.0]])
# 3 classes
class_indices = tf.constant([1., 2., 3.])
exp_ious = [[0.66666669, 0., 0.02255639, 0.15384616],
[0., 0.25, 0.00250627, 0.],
[0.015, 0.01, 0.90692127, 0.015]]
exp_max_ious = np.array([0.66666669, 0.25, 0.90692127, 0.15384616])
exp_max_indices = np.array([0, 1, 2, 0])
exp_pos_mask = np.array([True, False, True, False])
exp_class_and_background_indices = np.array([1, 0, 3, 0])
# Convert to box_list format
boxes_pred = box_list.BoxList(corners_pred)
boxes_gt = box_list.BoxList(corners_gt)
# Calculate IoU
iou = box_list_ops.iou(boxes_gt,
boxes_pred)
# Get max IoU, the dimension should match the anchors we are
# evaluating
max_ious = tf.reduce_max(iou, axis=0)
max_iou_indices = tf.argmax(iou, axis=0)
# Sample a mini-batch from anchors with highest IoU match
mini_batch_size = 4
# Custom positive/negative iou ranges
neg_2d_iou_range = [0.0, 0.3]
pos_2d_iou_range = [0.6, 0.7]
mb_mask, mb_pos_mask = \
self.mb_utils.sample_mini_batch(max_ious,
mini_batch_size,
neg_2d_iou_range,
pos_2d_iou_range)
mb_class_indices = self.mb_utils.mask_class_label_indices(
mb_pos_mask, mb_mask, max_iou_indices, class_indices)
with self.test_session() as sess:
iou_out = sess.run(iou)
max_ious_out, max_iou_indices_out = sess.run([max_ious,
max_iou_indices])
mb_mask_out, mb_pos_mask_out = sess.run([mb_mask,
mb_pos_mask])
class_indices_out = sess.run(mb_class_indices)
self.assertAllClose(iou_out, exp_ious)
self.assertAllClose(max_ious_out, exp_max_ious)
self.assertAllEqual(max_iou_indices_out, exp_max_indices)
self.assertAllEqual(exp_pos_mask, mb_pos_mask_out)
self.assertAllEqual(class_indices_out,
exp_class_and_background_indices)