Object Detection API: ssd_random_crop_pad_fixed_aspect_ratio incorrect use of min and max_padded_size_ratio

System information

• What is the top-level directory of the model you are using: /home//models/research/object_detection
• Have I written custom code (as opposed to using a stock example script provided in TensorFlow): No
• OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Ubuntnu 18.04
• TensorFlow installed from (source or binary): binary, pip
• TensorFlow version (use command below): 1.14
• Bazel version (if compiling from source):
• CUDA/cuDNN version:
• GPU model and memory:
• Exact command to reproduce:

Describe the problem

The documentation for the augmentation ssd_random_crop_pad_fixed_aspect_ratio says that

min_padded_size_ratio: min ratio of padded image height and width to the input image's height and width. max_padded_size_ratio: max ratio of padded image height and width to the input image's height and width.

However, looking at the code in core/preprocessor.py line 3655, the function random_pad_to_aspect_ratio is called on the cropped image. The final result is then incorrect.

Source code / logs

You can run this code from the object_detection directory assuming that lena is your test image. The final result should be 5 times bigger than the original but it is not.

from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

import functools
import os
import cv2
from absl.testing import parameterized

import numpy as np
import tensorflow as tf
from scipy.misc import imsave, imread

from object_detection import inputs
from object_detection.core import preprocessor
from object_detection.core import standard_fields as fields
from object_detection.utils import config_util
from object_detection.utils import test_case

FLAGS = tf.flags.FLAGS
tf.disable_eager_execution()
class DataAugmentationFnTest(test_case.TestCase):

  def test_apply_image_and_box_augmentation(self):
    data_augmentation_options = [
        (preprocessor.ssd_random_crop_pad_fixed_aspect_ratio, {
			    'min_object_covered': [1.0],
                            'aspect_ratio': 1.0,
                            'aspect_ratio_range': [(1.0, 1.0)],
                            'area_range': [(0.1, 1.0)],
                            'overlap_thresh': [1.0],
                            'clip_boxes': [False],
                            'random_coef': [0.0],
                            'min_padded_size_ratio': (5.0, 5.0),
                            'max_padded_size_ratio': (5.0, 5.0)})
    ]
    data_augmentation_fn = functools.partial(
        inputs.augment_input_data,
        data_augmentation_options=data_augmentation_options)
    tensor_dict = {
        fields.InputDataFields.image:
            tf.constant(imread('lena.png').astype(np.float32)),
        fields.InputDataFields.groundtruth_boxes:
            tf.constant(np.array([[.5, .5, .51, .51]], np.float32)),
        fields.InputDataFields.groundtruth_classes:
            tf.constant(np.array([1.0], np.float32))
    }
    augmented_tensor_dict = data_augmentation_fn(tensor_dict=tensor_dict)
    with self.session() as sess:
      augmented_tensor_dict_out = sess.run(augmented_tensor_dict)
    print("Final Shape" + augmented_tensor_dict_out[fields.InputDataFields.image].shape)

if __name__ == '__main__':
  tf.test.main()