[feat] [assistant] [#I5EWHP] add new video operator VerticalFlip

docs/api/api_python/dataset_vision/mindspore.dataset.vision.VerticalFlip.rst

@@ -6,4 +6,4 @@ mindspore.dataset.vision.VerticalFlip
     对输入图像进行垂直翻转。
 
     异常：
-        - **RuntimeError** - 如果输入的Tensor不是 <H, W> 或 <H, W, C> 格式。
+        - **RuntimeError** - 如果输入图像的shape不是 <H, W> 或 <..., H, W, C>。

mindspore/ccsrc/minddata/dataset/kernels/image/vertical_flip_op.cc

@@ -1,5 +1,5 @@
 /**
- * Copyright 2021 Huawei Technologies Co., Ltd
+ * Copyright 2021-2022 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.
@@ -16,17 +16,39 @@
 
 #include "minddata/dataset/kernels/image/vertical_flip_op.h"
 
+#include "minddata/dataset/kernels/data/data_utils.h"
 #include "minddata/dataset/kernels/image/image_utils.h"
 
 namespace mindspore {
 namespace dataset {
 Status VerticalFlipOp::Compute(const std::shared_ptr<Tensor> &input, std::shared_ptr<Tensor> *output) {
   IO_CHECK(input, output);
+  RETURN_IF_NOT_OK(ValidateImage(input, "VerticalFlip", {1, 2, 3, 4, 5, 6, 10, 11, 12}));
+  dsize_t rank = input->shape().Rank();
+  if (rank <= kDefaultImageRank) {
+    // [H, W] or [H, W, C]
+    RETURN_IF_NOT_OK(VerticalFlip(input, output));
+  } else {
+    // reshape [..., H, W, C] to [N, H, W, C]
+    auto input_shape = input->shape();
+    dsize_t num_batch = input->Size() / (input_shape[-3] * input_shape[-2] * input_shape[-1]);
+    TensorShape new_shape({num_batch, input_shape[-3], input_shape[-2], input_shape[-1]});
+    RETURN_IF_NOT_OK(input->Reshape(new_shape));
 
-  RETURN_IF_NOT_OK(ValidateImageDtype("VerticalFlip", input->type()));
-  RETURN_IF_NOT_OK(ValidateImageRank("VerticalFlip", input->Rank()));
+    // split [N, H, W, C] to N [H, W, C], and vertical flip N [H, W, C]
+    std::vector<std::shared_ptr<Tensor>> input_vector_hwc, output_vector_hwc;
+    RETURN_IF_NOT_OK(BatchTensorToTensorVector(input, &input_vector_hwc));
+    for (int i = 0; i < num_batch; i++) {
+      std::shared_ptr<Tensor> flip;
+      RETURN_IF_NOT_OK(VerticalFlip(input_vector_hwc[i], &flip));
+      output_vector_hwc.push_back(flip);
+    }
 
-  return VerticalFlip(input, output);
+    // integrate N [H, W, C] to [N, H, W, C], and reshape [..., H, W, C]
+    RETURN_IF_NOT_OK(TensorVectorToBatchTensor(output_vector_hwc, output));
+    RETURN_IF_NOT_OK((*output)->Reshape(input_shape));
+  }
+  return Status::OK();
 }
 }  // namespace dataset
 }  // namespace mindspore

mindspore/python/mindspore/dataset/vision/c_transforms.py

@@ -2611,7 +2611,7 @@ class VerticalFlip(ImageTensorOperation):
     Flip the input image vertically.
 
     Raises:
-        RuntimeError: If given tensor shape is not <H, W> or <H, W, C>.
+        RuntimeError: If given tensor shape is not <H, W> or <..., H, W, C>.
 
     Supported Platforms:
         ``CPU``

mindspore/python/mindspore/dataset/vision/transforms.py

@@ -4097,7 +4097,7 @@ class VerticalFlip(ImageTensorOperation):
     Flip the input image vertically.
 
     Raises:
-        RuntimeError: If given tensor shape is not <H, W> or <H, W, C>.
+        RuntimeError: If given tensor shape is not <H, W> or <..., H, W, C>.
 
     Supported Platforms:
         ``CPU``

tests/ut/cpp/dataset/c_api_vision_vertical_flip_test.cc

@@ -89,3 +89,54 @@ TEST_F(MindDataTestVerticalFlip, TestVerticalFlipEager) {
 
   EXPECT_EQ(rc, Status::OK());
 }
+
+/// Feature: VerticalFlip op
+/// Description: Test VerticalFlip op by processing tensor with dim more than 3
+/// Expectation: Output is equal to the expected output
+TEST_F(MindDataTestVerticalFlip, TestVerticalFlipBatch) {
+  MS_LOG(INFO) << "Doing MindDataTestVerticalFlip-TestVerticalFlipBatch.";
+
+  // Create an ImageFolder Dataset
+  std::string folder_path = datasets_root_path_ + "/testPK/data/";
+  std::shared_ptr<Dataset> ds = ImageFolder(folder_path, true, std::make_shared<RandomSampler>(false, 5));
+  EXPECT_NE(ds, nullptr);
+
+  // Create a Repeat operation on ds
+  int32_t repeat_num = 3;
+  ds = ds->Repeat(repeat_num);
+  EXPECT_NE(ds, nullptr);
+
+  // Create a Batch operation on ds, choose batch size 3 to test high dimension input
+  int32_t batch_size = 3;
+  ds = ds->Batch(batch_size);
+  EXPECT_NE(ds, nullptr);
+
+  // Create objects for the tensor ops
+  auto vertical_flip = std::make_shared<vision::VerticalFlip>();
+
+  // Create a Map operation on ds
+  ds = ds->Map({vertical_flip});
+  EXPECT_NE(ds, nullptr);
+
+  // Create an iterator over the result of the above dataset
+  // This will trigger the creation of the Execution Tree and launch it.
+  std::shared_ptr<Iterator> iter = ds->CreateIterator();
+  EXPECT_NE(iter, nullptr);
+
+  // Iterate the dataset and get each row
+  std::unordered_map<std::string, mindspore::MSTensor> row;
+  ASSERT_OK(iter->GetNextRow(&row));
+
+  uint64_t i = 0;
+  while (row.size() != 0) {
+    i++;
+    auto image = row["image"];
+    MS_LOG(INFO) << "Tensor image shape: " << image.Shape();
+    ASSERT_OK(iter->GetNextRow(&row));
+  }
+
+  EXPECT_EQ(i, 5);
+
+  // Manually terminate the pipeline
+  iter->Stop();
+}

tests/ut/python/dataset/test_vertical_flip.py

@@ -16,6 +16,7 @@
 Testing VerticalFlip Python API
 """
 import cv2
+import numpy as np
 
 import mindspore.dataset as ds
 import mindspore.dataset.vision as vision
@@ -26,6 +27,14 @@ from util import visualize_image, diff_mse
 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"
 IMAGE_FILE = "../data/dataset/apple.jpg"
+FOUR_DIM_DATA = [[[[1, 2, 3], [3, 4, 3]], [[5, 6, 3], [7, 8, 3]]],
+                 [[[9, 10, 3], [11, 12, 3]], [[13, 14, 3], [15, 16, 3]]]]
+FIVE_DIM_DATA = [[[[[1, 2, 3], [3, 4, 3]], [[5, 6, 3], [7, 8, 3]]],
+                  [[[9, 10, 3], [11, 12, 3]], [[13, 14, 3], [15, 16, 3]]]]]
+FOUR_DIM_RES = [[[[5, 6, 3], [7, 8, 3]], [[1, 2, 3], [3, 4, 3]]],
+                [[[13, 14, 3], [15, 16, 3]], [[9, 10, 3], [11, 12, 3]]]]
+FIVE_DIM_RES = [[[[[5, 6, 3], [7, 8, 3]], [[1, 2, 3], [3, 4, 3]]],
+                 [[[13, 14, 3], [15, 16, 3]], [[9, 10, 3], [11, 12, 3]]]]]
 
 
 def test_vertical_flip_pipeline(plot=False):
@@ -78,6 +87,111 @@ def test_vertical_flip_eager():
     assert mse == 0
 
 
+def test_vertical_flip_video_op_1d():
+    """
+    Feature: VerticalFlip op
+    Description: Test VerticalFlip op by processing tensor with dim 1
+    Expectation:  Error is raised as expected
+    """
+    logger.info("Test VerticalFlip with 1 dimension input")
+    data = [1]
+    input_mindspore = np.array(data).astype(np.uint8)
+    vertical_flip_op = vision.VerticalFlip()
+    try:
+        vertical_flip_op(input_mindspore)
+    except RuntimeError as e:
+        logger.info("Got an exception in DE: {}".format(str(e)))
+        assert "VerticalFlip: the image tensor should have at least two dimensions. You may need to perform " \
+               "Decode first." in str(e)
+
+
+def test_vertical_flip_video_op_4d():
+    """
+    Feature: VerticalFlip op
+    Description: Test VerticalFlip op by processing tensor with dim more than 3 (dim 4)
+    Expectation: The dataset is processed successfully
+    """
+    logger.info("Test VerticalFlip with 4 dimension input")
+    input_4_dim = np.array(FOUR_DIM_DATA).astype(np.uint8)
+    input_4_shape = input_4_dim.shape
+    num_batch = input_4_shape[0]
+    out_4_list = []
+    batch_1d = 0
+    while batch_1d < num_batch:
+        out_4_list.append(cv2.flip(input_4_dim[batch_1d], 0))
+        batch_1d += 1
+    out_4_cv = np.array(out_4_list).astype(np.uint8)
+    vertical_flip_op = vision.VerticalFlip()
+    out_4_mindspore = vertical_flip_op(input_4_dim)
+
+    mse = diff_mse(out_4_mindspore, out_4_cv)
+    assert mse < 0.001
+
+
+def test_vertical_flip_video_op_5d():
+    """
+    Feature: VerticalFlip op
+    Description: process tensor with dim more than 3 (dim 5)
+    Expectation: process successfully
+    """
+    input_5_dim = np.array(FIVE_DIM_DATA).astype(np.uint8)
+    input_5_shape = input_5_dim.shape
+    num_batch_1d = input_5_shape[0]
+    num_batch_2d = input_5_shape[1]
+    out_5_list = []
+    batch_1d = 0
+    batch_2d = 0
+    while batch_1d < num_batch_1d:
+        while batch_2d < num_batch_2d:
+            out_5_list.append(cv2.flip(input_5_dim[batch_1d][batch_2d], 0))
+            batch_2d += 1
+        batch_1d += 1
+    out_5_cv = np.array(out_5_list).astype(np.uint8)
+    vertical_flip_op = vision.VerticalFlip()
+    out_5_mindspore = vertical_flip_op(input_5_dim)
+
+    mse = diff_mse(out_5_mindspore, out_5_cv)
+    assert mse < 0.001
+
+
+def test_vertical_flip_video_op_precision_eager():
+    """
+    Feature: VerticalFlip op
+    Description: Test VerticalFlip op by processing tensor with dim more than 3 (dim 4) in eager mode
+    Expectation: The dataset is processed successfully
+    """
+    logger.info("Test VerticalFlip eager with 4 dimension input")
+    input_mindspore = np.array(FOUR_DIM_DATA).astype(np.uint8)
+
+    vertical_flip_op = vision.VerticalFlip()
+    out_mindspore = vertical_flip_op(input_mindspore)
+    mse = diff_mse(out_mindspore, np.array(FOUR_DIM_RES).astype(np.uint8))
+    assert mse < 0.001
+
+
+def test_vertical_flip_video_op_precision_pipeline():
+    """
+    Feature: VerticalFlip op
+    Description: Test VerticalFlip op by processing tensor with dim more than 3 (dim 5) in pipeline mode
+    Expectation: The dataset is processed successfully
+    """
+    logger.info("Test VerticalFlip pipeline with 5 dimension input")
+    data = np.array(FIVE_DIM_DATA).astype(np.uint8)
+    expand_data = np.expand_dims(data, axis=0)
+
+    dataset = ds.NumpySlicesDataset(expand_data, column_names=["col1"], shuffle=False)
+    vertical_flip_op = vision.VerticalFlip()
+    dataset = dataset.map(operations=vertical_flip_op, input_columns=["col1"])
+    for item in dataset.create_dict_iterator(output_numpy=True):
+        mse = diff_mse(item["col1"], np.array(FIVE_DIM_RES).astype(np.uint8))
+        assert mse < 0.001
+
+
 if __name__ == "__main__":
     test_vertical_flip_pipeline(plot=False)
     test_vertical_flip_eager()
+    test_vertical_flip_video_op_1d()
+    test_vertical_flip_video_op_4d()
+    test_vertical_flip_video_op_5d()
+    test_vertical_flip_video_op_precision_eager()
+    test_vertical_flip_video_op_precision_pipeline()