self code-review to clean up

research/object_detection/box_coders/detr_box_coder.py

-# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
+# Copyright 2020 The TensorFlow Authors. All Rights Reserved.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.

research/object_detection/core/losses.py

@@ -225,7 +225,7 @@ class WeightedGIOULocalizationLoss(Loss):
 
     Args:
       prediction_tensor: A float tensor of shape [batch_size, num_anchors, 4]
-        representing the decoded predicted boxes
+        representing the predicted boxes in the form
       target_tensor: A float tensor of shape [batch_size, num_anchors, 4]
         representing the decoded target boxes
       weights: a float tensor of shape [batch_size, num_anchors]
@@ -236,10 +236,8 @@ class WeightedGIOULocalizationLoss(Loss):
     """
     batch_size, num_anchors, _ = shape_utils.combined_static_and_dynamic_shape(
         prediction_tensor)
-    predicted_boxes = ops.cy_cx_h_w_to_ymin_xmin_ymax_xmax_coords(
-        tf.reshape(prediction_tensor, [-1, 4]))
-    target_boxes = ops.cy_cx_h_w_to_ymin_xmin_ymax_xmax_coords(
-        tf.reshape(target_tensor, [-1, 4]))
+    predicted_boxes = tf.reshape(prediction_tensor, [-1, 4])
+    target_boxes = tf.reshape(target_tensor, [-1, 4])
 
     per_anchor_iou_loss = 1 - ops.giou(predicted_boxes, target_boxes)
     return tf.reshape(tf.reshape(weights, [-1]) * per_anchor_iou_loss,

research/object_detection/core/losses_test.py

@@ -28,6 +28,7 @@ from object_detection.core import box_list
 from object_detection.core import losses
 from object_detection.core import matcher
 from object_detection.utils import test_case
+from object_detection.utils import ops
 
 
 class WeightedL2LocalizationLossTest(test_case.TestCase):
@@ -206,10 +207,12 @@ class WeightedGIOULocalizationLossTest(test_case.TestCase):
                                         [0, 0, 0, 0]]])
       target_tensor = tf.constant([[[1.5, 0, 2.4, 1],
                                     [0, 0, 1, 1],
-                                    [7.5, 7.5, 2.5, 2.5]]])
+                                    [5, 5, 10, 10]]])
       weights = [[1.0, .5, 2.0]]
       loss_op = losses.WeightedGIOULocalizationLoss()
-      loss = loss_op(prediction_tensor, target_tensor, weights=weights)
+      loss = loss_op(prediction_tensor,
+                     target_tensor,
+                     weights=weights)
       loss = tf.reduce_sum(loss)
       return loss
     exp_loss = 3.5

research/object_detection/core/region_similarity_calculator.py

@@ -34,7 +34,8 @@ from object_detection.core import standard_fields as fields
 class RegionSimilarityCalculator(six.with_metaclass(ABCMeta, object)):
   """Abstract base class for region similarity calculator."""
 
-  def compare(self, boxlist1, boxlist2, scope=None, groundtruth_labels=None, predicted_labels=None):
+  def compare(self, boxlist1, boxlist2, scope=None,
+              groundtruth_labels=None, predicted_labels=None):
     """Computes matrix of pairwise similarity between BoxLists.
 
     This op (to be overridden) computes a measure of pairwise similarity between
@@ -59,7 +60,8 @@ class RegionSimilarityCalculator(six.with_metaclass(ABCMeta, object)):
       return self._compare(boxlist1, boxlist2, groundtruth_labels, predicted_labels)
 
   @abstractmethod
-  def _compare(self, boxlist1, boxlist2, groundtruth_labels=None, predicted_labels=None):
+  def _compare(self, boxlist1, boxlist2,
+               groundtruth_labels=None, predicted_labels=None):
     pass
 
 

research/object_detection/core/region_similarity_calculator_test.py

@@ -102,7 +102,8 @@ class RegionSimilarityCalculatorTest(test_case.TestCase):
       boxes1 = box_list.BoxList(corners1)
       boxes2 = box_list.BoxList(corners2)
       detr_similarity_calculator = region_similarity_calculator.DETRSimilarity()
-      detr_similarity = detr_similarity_calculator.compare(boxes1, boxes2, None, groundtruth_labels, predicted_labels)
+      detr_similarity = detr_similarity_calculator.compare(
+          boxes1, boxes2, None, groundtruth_labels, predicted_labels)
       return detr_similarity
     exp_output = [[2.0, -2.0/3.0 + 1.0 - 20.0]]
     sim_output = self.execute(graph_fn, [])

research/object_detection/core/target_assigner.py

@@ -142,8 +142,8 @@ class TargetAssigner(object):
         aware of groundtruth weights. Additionally, `cls_weights` and
         `reg_weights` are calculated using groundtruth weights as an added
         safety.
-      class_predictions: a float tensor of shape [N, num_classes] containing class
-        predictions from the model, to be used in certain matchers.
+      class_predictions: A tensor with shape [max_num_boxes, d_1, d_2, ..., d_k]
+        to be used by certain similarity calculators.
 
     Returns:
       cls_targets: a float32 tensor with shape [num_anchors, d_1, d_2 ... d_k],
@@ -200,7 +200,8 @@ class TargetAssigner(object):
     with tf.control_dependencies(
         [unmatched_shape_assert, labels_and_box_shapes_assert]):
       
-      match_quality_matrix = self._similarity_calc.compare(groundtruth_boxes,
+      match_quality_matrix = self._similarity_calc.compare(
+          groundtruth_boxes,
           anchors,
           groundtruth_labels=groundtruth_labels,
           predicted_labels=class_predictions)
@@ -481,7 +482,8 @@ def batch_assign(target_assigner,
       function (which have shape [num_gt_boxes, d_1, d_2, ..., d_k]).
     gt_weights_batch: A list of 1-D tf.float32 tensors of shape
       [num_boxes] containing weights for groundtruth boxes.
-    class_predictions: A 
+    class_predictions: A tensor with shape [batch_size, max_num_boxes,
+      d_1, d_2, ..., d_k] to be used by certain similarity calculators.
 
   Returns:
     batch_cls_targets: a tensor with shape [batch_size, num_anchors,
@@ -531,8 +533,8 @@ def batch_assign(target_assigner,
       class_predictions):
     (cls_targets, cls_weights,
      reg_targets, reg_weights, match) = target_assigner.assign(
-         anchors, gt_boxes, gt_class_targets, unmatched_class_label, gt_weights,
-         class_preds)
+         anchors, gt_boxes, gt_class_targets, unmatched_class_label,
+         gt_weights, class_preds)
     cls_targets_list.append(cls_targets)
     cls_weights_list.append(cls_weights)
     reg_targets_list.append(reg_targets)

research/object_detection/utils/ops.py

@@ -1185,7 +1185,7 @@ def giou(boxes1, boxes2):
 
   return shape_utils.static_or_dynamic_map_fn(two_boxes_giou, [boxes1, boxes2])
 
-def cy_cx_h_w_to_ymin_xmin_ymax_xmax_coords(input_tensor):
+def cy_cx_h_w_to_ymin_xmin_ymax_xmax(input_tensor):
   """Converts input boxes from center to corner representation."""
   reshaped_encodings = tf.reshape(input_tensor, [-1, 4])
   ycenter = tf.gather(reshaped_encodings, [0], axis=1)

research/object_detection/utils/ops_test.py

@@ -1664,7 +1664,7 @@ class TestCoordinateConversion(test_case.TestCase):
     def graph_fn():
       boxes = tf.constant([[3, 3, 5, 5], [3, 4, 2, 6]], dtype=tf.float32)
 
-      converted = ops.cy_cx_h_w_to_ymin_xmin_ymax_xmax_coords(boxes)
+      converted = ops.cy_cx_h_w_to_ymin_xmin_ymax_xmax(boxes)
       self.assertEqual(converted.dtype, tf.float32)
 
       return converted