forked from mindspore-Ecosystem/mindspore
187 lines
6.9 KiB
Python
187 lines
6.9 KiB
Python
# Copyright 2020-2022 Huawei Technologies Co., Ltd
|
|
#
|
|
# Licensed under the Apache License, Version 2.0 (the "License");
|
|
# you may not use this file except in compliance with the License.
|
|
# You may obtain a copy of the License at
|
|
#
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
#
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
# See the License for the specific language governing permissions and
|
|
# limitations under the License.
|
|
# ==============================================================================
|
|
"""
|
|
Testing HWC2CHW op in DE
|
|
"""
|
|
import numpy as np
|
|
import pytest
|
|
import mindspore.dataset as ds
|
|
import mindspore.dataset.transforms.transforms
|
|
import mindspore.dataset.vision.transforms as vision
|
|
from mindspore import log as logger
|
|
from util import diff_mse, visualize_list, save_and_check_md5
|
|
|
|
GENERATE_GOLDEN = False
|
|
|
|
DATA_DIR = ["../data/dataset/test_tf_file_3_images/train-0000-of-0001.data"]
|
|
SCHEMA_DIR = "../data/dataset/test_tf_file_3_images/datasetSchema.json"
|
|
|
|
|
|
def test_hwc2chw_callable():
|
|
"""
|
|
Feature: HWC2CHW op
|
|
Description: Test HWC2CHW op is callable.
|
|
Expectation: Valid input succeeds. Invalid input fails.
|
|
"""
|
|
logger.info("Test HWC2CHW callable")
|
|
img = np.zeros([50, 50, 3])
|
|
assert img.shape == (50, 50, 3)
|
|
|
|
# test one tensor
|
|
img1 = vision.HWC2CHW()(img)
|
|
assert img1.shape == (3, 50, 50)
|
|
|
|
# test one tensor with 5 channels
|
|
img2 = np.zeros([50, 50, 5])
|
|
assert img2.shape == (50, 50, 5)
|
|
img3 = vision.HWC2CHW()(img2)
|
|
assert img3.shape == (5, 50, 50)
|
|
|
|
# test input multiple tensors
|
|
with pytest.raises(RuntimeError) as info:
|
|
imgs = [img, img]
|
|
_ = vision.HWC2CHW()(*imgs)
|
|
assert "The op is OneToOne, can only accept one tensor as input" in str(info.value)
|
|
|
|
with pytest.raises(RuntimeError) as info:
|
|
_ = vision.HWC2CHW()(img, img)
|
|
assert "The op is OneToOne, can only accept one tensor as input" in str(info.value)
|
|
|
|
|
|
def test_hwc2chw_multi_channels():
|
|
"""
|
|
Feature: Test HWC2CHW feature
|
|
Description: The input is a HWC format array with 5 channels
|
|
Expectation: success
|
|
"""
|
|
logger.info("Test HWC2CHW with data of 5 channels")
|
|
|
|
# create numpy array in HWC format with shape (4, 2, 5) like a fake image with 5 channels
|
|
raw_data = np.random.rand(4, 2, 5).astype(np.float32)
|
|
expect_output = np.transpose(raw_data, (2, 0, 1))
|
|
|
|
# NumpySliceDataset support accept data stored in list, tuple etc, here only one row data in list.
|
|
input_data = np.array([raw_data])
|
|
dataset = ds.NumpySlicesDataset(input_data, column_names=["col1"], shuffle=False)
|
|
|
|
hwc2chw = vision.HWC2CHW()
|
|
dataset = dataset.map(hwc2chw, input_columns=["col1"])
|
|
for item in dataset.create_tuple_iterator(output_numpy=True):
|
|
assert np.allclose(item[0], expect_output)
|
|
|
|
|
|
def test_hwc2chw(plot=False):
|
|
"""
|
|
Feature: HWC2CHW op
|
|
Description: Test HWC2CHW op in pipeline
|
|
Expectation: Pipelines succeed with comparison mse=0
|
|
"""
|
|
logger.info("Test HWC2CHW")
|
|
|
|
# First dataset
|
|
data1 = ds.TFRecordDataset(DATA_DIR, SCHEMA_DIR, columns_list=["image"], shuffle=False)
|
|
decode_op = vision.Decode()
|
|
hwc2chw_op = vision.HWC2CHW()
|
|
data1 = data1.map(operations=decode_op, input_columns=["image"])
|
|
data1 = data1.map(operations=hwc2chw_op, input_columns=["image"])
|
|
|
|
# Second dataset
|
|
data2 = ds.TFRecordDataset(DATA_DIR, SCHEMA_DIR, columns_list=["image"], shuffle=False)
|
|
data2 = data2.map(operations=decode_op, input_columns=["image"])
|
|
|
|
image_transposed = []
|
|
image = []
|
|
for item1, item2 in zip(data1.create_dict_iterator(num_epochs=1, output_numpy=True),
|
|
data2.create_dict_iterator(num_epochs=1, output_numpy=True)):
|
|
transposed_item = item1["image"].copy()
|
|
original_item = item2["image"].copy()
|
|
image_transposed.append(transposed_item.transpose(1, 2, 0))
|
|
image.append(original_item)
|
|
|
|
# check if the shape of data is transposed correctly
|
|
# transpose the original image from shape (H,W,C) to (C,H,W)
|
|
mse = diff_mse(transposed_item, original_item.transpose(2, 0, 1))
|
|
assert mse == 0
|
|
if plot:
|
|
visualize_list(image, image_transposed)
|
|
|
|
|
|
def test_hwc2chw_md5():
|
|
"""
|
|
Feature: HWC2CHW op
|
|
Description: Test HWC2CHW op with md5 check.
|
|
Expectation: Pipeline results match in md5 comparison
|
|
"""
|
|
logger.info("Test HWC2CHW with md5 comparison")
|
|
|
|
# First dataset
|
|
data1 = ds.TFRecordDataset(DATA_DIR, SCHEMA_DIR, columns_list=["image"], shuffle=False)
|
|
decode_op = vision.Decode()
|
|
hwc2chw_op = vision.HWC2CHW()
|
|
data1 = data1.map(operations=decode_op, input_columns=["image"])
|
|
data1 = data1.map(operations=hwc2chw_op, input_columns=["image"])
|
|
|
|
# Compare with expected md5 from images
|
|
filename = "HWC2CHW_01_result.npz"
|
|
save_and_check_md5(data1, filename, generate_golden=GENERATE_GOLDEN)
|
|
|
|
|
|
def test_hwc2chw_comp(plot=False):
|
|
"""
|
|
Feature: HWC2CHW op
|
|
Description: Test HWC2CHW between Python and C image augmentation
|
|
Expectation: Image augmentations should be almost the same with mse < 0.001
|
|
"""
|
|
logger.info("Test HWC2CHW with C and Python image augmentation comparison")
|
|
|
|
# First dataset
|
|
data1 = ds.TFRecordDataset(DATA_DIR, SCHEMA_DIR, columns_list=["image"], shuffle=False)
|
|
decode_op = vision.Decode()
|
|
hwc2chw_op = vision.HWC2CHW()
|
|
data1 = data1.map(operations=decode_op, input_columns=["image"])
|
|
data1 = data1.map(operations=hwc2chw_op, input_columns=["image"])
|
|
|
|
# Second dataset
|
|
data2 = ds.TFRecordDataset(DATA_DIR, SCHEMA_DIR, columns_list=["image"], shuffle=False)
|
|
transforms = [
|
|
vision.Decode(True),
|
|
vision.ToTensor()]
|
|
transform = mindspore.dataset.transforms.transforms.Compose(transforms)
|
|
data2 = data2.map(operations=transform, input_columns=["image"])
|
|
|
|
image_c_transposed = []
|
|
image_py_transposed = []
|
|
for item1, item2 in zip(data1.create_dict_iterator(num_epochs=1, output_numpy=True),
|
|
data2.create_dict_iterator(num_epochs=1, output_numpy=True)):
|
|
c_image = item1["image"]
|
|
py_image = (item2["image"] * 255).astype(np.uint8)
|
|
|
|
# Compare images between that applying C implementation and Python implementation image augmentations
|
|
mse = diff_mse(py_image, c_image)
|
|
# Note: The images aren't exactly the same due to rounding error
|
|
assert mse < 0.001
|
|
if plot:
|
|
image_c_transposed.append(c_image.transpose(1, 2, 0))
|
|
image_py_transposed.append(py_image.transpose(1, 2, 0))
|
|
visualize_list(image_c_transposed, image_py_transposed, visualize_mode=2)
|
|
|
|
|
|
if __name__ == '__main__':
|
|
test_hwc2chw_callable()
|
|
test_hwc2chw_multi_channels()
|
|
test_hwc2chw(True)
|
|
test_hwc2chw_md5()
|
|
test_hwc2chw_comp(True)
|