!4621 [MD] change api for random solarized_op

Merge pull request !4621 from nhussain/add_solarize_cpp_tests

mindspore/ccsrc/minddata/dataset/api/python/bindings/dataset/kernels/image/bindings.cc

@@ -422,7 +422,7 @@ PYBIND_REGISTER(
 PYBIND_REGISTER(RandomSolarizeOp, 1, ([](const py::module *m) {
                   (void)py::class_<RandomSolarizeOp, TensorOp, std::shared_ptr<RandomSolarizeOp>>(*m,
                                                                                                   "RandomSolarizeOp")
-                    .def(py::init<uint8_t, uint8_t>());
+                    .def(py::init<std::vector<uint8_t>>());
                 }));
 
 }  // namespace dataset

mindspore/ccsrc/minddata/dataset/api/transforms.cc

@@ -241,8 +241,8 @@ std::shared_ptr<RandomRotationOperation> RandomRotation(std::vector<float> degre
 }
 
 // Function to create RandomSolarizeOperation.
-std::shared_ptr<RandomSolarizeOperation> RandomSolarize(uint8_t threshold_min, uint8_t threshold_max) {
-  auto op = std::make_shared<RandomSolarizeOperation>(threshold_min, threshold_max);
+std::shared_ptr<RandomSolarizeOperation> RandomSolarize(std::vector<uint8_t> threshold) {
+  auto op = std::make_shared<RandomSolarizeOperation>(threshold);
   // Input validation
   if (!op->ValidateParams()) {
     return nullptr;
@@ -811,19 +811,22 @@ std::shared_ptr<TensorOp> RandomSharpnessOperation::Build() {
 }
 
 // RandomSolarizeOperation.
-RandomSolarizeOperation::RandomSolarizeOperation(uint8_t threshold_min, uint8_t threshold_max)
-    : threshold_min_(threshold_min), threshold_max_(threshold_max) {}
+RandomSolarizeOperation::RandomSolarizeOperation(std::vector<uint8_t> threshold) : threshold_(threshold) {}
 
 bool RandomSolarizeOperation::ValidateParams() {
-  if (threshold_max_ < threshold_min_) {
-    MS_LOG(ERROR) << "RandomSolarize: threshold_max must be greater or equal to threshold_min";
+  if (threshold_.size() != 2) {
+    MS_LOG(ERROR) << "RandomSolarize: threshold vector has incorrect size: " << threshold_.size();
+    return false;
+  }
+  if (threshold_.at(0) > threshold_.at(1)) {
+    MS_LOG(ERROR) << "RandomSolarize: threshold must be passed in a min, max format";
     return false;
   }
   return true;
 }
 
 std::shared_ptr<TensorOp> RandomSolarizeOperation::Build() {
-  std::shared_ptr<RandomSolarizeOp> tensor_op = std::make_shared<RandomSolarizeOp>(threshold_min_, threshold_max_);
+  std::shared_ptr<RandomSolarizeOp> tensor_op = std::make_shared<RandomSolarizeOp>(threshold_);
   return tensor_op;
 }
 

mindspore/ccsrc/minddata/dataset/include/transforms.h

@@ -249,10 +249,9 @@ std::shared_ptr<RandomSharpnessOperation> RandomSharpness(std::vector<float> deg
 
 /// \brief Function to create a RandomSolarize TensorOperation.
 /// \notes Invert pixels within specified range. If min=max, then it inverts all pixel above that threshold
-/// \param[in] threshold_min - lower limit
-/// \param[in] threshold_max - upper limit
+/// \param[in] threshold - a vector with two elements specifying the pixel range to invert.
 /// \return Shared pointer to the current TensorOperation.
-std::shared_ptr<RandomSolarizeOperation> RandomSolarize(uint8_t threshold_min = 0, uint8_t threshold_max = 255);
+std::shared_ptr<RandomSolarizeOperation> RandomSolarize(std::vector<uint8_t> threshold = {0, 255});
 
 /// \brief Function to create a RandomVerticalFlip TensorOperation.
 /// \notes Tensor operation to perform random vertical flip.
@@ -657,7 +656,7 @@ class SwapRedBlueOperation : public TensorOperation {
 
 class RandomSolarizeOperation : public TensorOperation {
  public:
-  explicit RandomSolarizeOperation(uint8_t threshold_min, uint8_t threshold_max);
+  explicit RandomSolarizeOperation(std::vector<uint8_t> threshold);
 
   ~RandomSolarizeOperation() = default;
 
@@ -666,8 +665,7 @@ class RandomSolarizeOperation : public TensorOperation {
   bool ValidateParams() override;
 
  private:
-  uint8_t threshold_min_;
-  uint8_t threshold_max_;
+  std::vector<uint8_t> threshold_;
 };
 }  // namespace vision
 }  // namespace api

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

@@ -13,6 +13,8 @@
  * See the License for the specific language governing permissions and
  * limitations under the License.
  */
+#include <vector>
+
 #include "minddata/dataset/kernels/image/random_solarize_op.h"
 #include "minddata/dataset/kernels/image/solarize_op.h"
 #include "minddata/dataset/kernels/image/image_utils.h"
@@ -24,6 +26,9 @@ namespace dataset {
 
 Status RandomSolarizeOp::Compute(const std::shared_ptr<Tensor> &input, std::shared_ptr<Tensor> *output) {
   IO_CHECK(input, output);
+
+  uint8_t threshold_min_ = threshold_[0], threshold_max_ = threshold_[1];
+
   CHECK_FAIL_RETURN_UNEXPECTED(threshold_min_ <= threshold_max_,
                                "threshold_min must be smaller or equal to threshold_max.");
 
@@ -35,7 +40,8 @@ Status RandomSolarizeOp::Compute(const std::shared_ptr<Tensor> &input, std::shar
     threshold_min = threshold_max;
     threshold_max = temp;
   }
-  std::unique_ptr<SolarizeOp> op(new SolarizeOp(threshold_min, threshold_max));
+  std::vector<uint8_t> inputs = {threshold_min, threshold_max};
+  std::unique_ptr<SolarizeOp> op(new SolarizeOp(inputs));
   return op->Compute(input, output);
 }
 }  // namespace dataset

mindspore/ccsrc/minddata/dataset/kernels/image/random_solarize_op.h

@@ -19,6 +19,7 @@
 
 #include <memory>
 #include <string>
+#include <vector>
 
 #include "minddata/dataset/core/tensor.h"
 #include "minddata/dataset/kernels/tensor_op.h"
@@ -31,10 +32,7 @@ namespace dataset {
 class RandomSolarizeOp : public SolarizeOp {
  public:
   // Pick a random threshold value to solarize the image with
-  explicit RandomSolarizeOp(uint8_t threshold_min = 0, uint8_t threshold_max = 255)
-      : threshold_min_(threshold_min), threshold_max_(threshold_max) {
-    rnd_.seed(GetSeed());
-  }
+  explicit RandomSolarizeOp(std::vector<uint8_t> threshold = {0, 255}) : threshold_(threshold) { rnd_.seed(GetSeed()); }
 
   ~RandomSolarizeOp() = default;
 
@@ -43,8 +41,7 @@ class RandomSolarizeOp : public SolarizeOp {
   std::string Name() const override { return kRandomSolarizeOp; }
 
  private:
-  uint8_t threshold_min_;
-  uint8_t threshold_max_;
+  std::vector<uint8_t> threshold_;
   std::mt19937 rnd_;
 };
 }  // namespace dataset

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

@@ -27,6 +27,8 @@ const uint8_t kPixelValue = 255;
 Status SolarizeOp::Compute(const std::shared_ptr<Tensor> &input, std::shared_ptr<Tensor> *output) {
   IO_CHECK(input, output);
 
+  uint8_t threshold_min_ = threshold_[0], threshold_max_ = threshold_[1];
+
   CHECK_FAIL_RETURN_UNEXPECTED(threshold_min_ <= threshold_max_,
                                "threshold_min must be smaller or equal to threshold_max.");
 

mindspore/ccsrc/minddata/dataset/kernels/image/solarize_op.h

@@ -19,6 +19,7 @@
 
 #include <memory>
 #include <string>
+#include <vector>
 
 #include "minddata/dataset/core/tensor.h"
 #include "minddata/dataset/kernels/tensor_op.h"
@@ -28,8 +29,7 @@ namespace mindspore {
 namespace dataset {
 class SolarizeOp : public TensorOp {
  public:
-  explicit SolarizeOp(uint8_t threshold_min = 0, uint8_t threshold_max = 255)
-      : threshold_min_(threshold_min), threshold_max_(threshold_max) {}
+  explicit SolarizeOp(std::vector<uint8_t> threshold = {0, 255}) : threshold_(threshold) {}
 
   ~SolarizeOp() = default;
 
@@ -38,8 +38,7 @@ class SolarizeOp : public TensorOp {
   std::string Name() const override { return kSolarizeOp; }
 
  private:
-  uint8_t threshold_min_;
-  uint8_t threshold_max_;
+  std::vector<uint8_t> threshold_;
 };
 }  // namespace dataset
 }  // namespace mindspore

mindspore/dataset/transforms/vision/c_transforms.py

@@ -1045,4 +1045,4 @@ class RandomSolarize(cde.RandomSolarizeOp):
 
     @check_random_solarize
     def __init__(self, threshold=(0, 255)):
-        super().__init__(*threshold)
+        super().__init__(threshold)

tests/ut/cpp/dataset/c_api_transforms_test.cc

@@ -1109,29 +1109,59 @@ TEST_F(MindDataTestPipeline, TestUniformAugWithOps) {
   iter->Stop();
 }
 
-TEST_F(MindDataTestPipeline, TestRandomSolarize) {
+TEST_F(MindDataTestPipeline, TestRandomSolarizeSucess1) {
+  MS_LOG(INFO) << "Doing MindDataTestPipeline-TestRandomSolarize.";
+
   // Create an ImageFolder Dataset
   std::string folder_path = datasets_root_path_ + "/testPK/data/";
   std::shared_ptr<Dataset> ds = ImageFolder(folder_path, true, RandomSampler(false, 10));
   EXPECT_NE(ds, nullptr);
 
-  // Create a Repeat operation on ds
-  int32_t repeat_num = 2;
-  ds = ds->Repeat(repeat_num);
-  EXPECT_NE(ds, nullptr);
-
   // Create objects for the tensor ops
-  std::shared_ptr<TensorOperation> random_solarize =
-    mindspore::dataset::api::vision::RandomSolarize(23, 23);  // vision::RandomSolarize();
+  std::vector<uint8_t> threshold = {10, 100};
+  std::shared_ptr<TensorOperation> random_solarize = mindspore::dataset::api::vision::RandomSolarize(threshold);
   EXPECT_NE(random_solarize, nullptr);
 
   // Create a Map operation on ds
   ds = ds->Map({random_solarize});
   EXPECT_NE(ds, nullptr);
 
-  // Create a Batch operation on ds
-  int32_t batch_size = 1;
-  ds = ds->Batch(batch_size);
+  // 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, std::shared_ptr<Tensor>> row;
+  iter->GetNextRow(&row);
+
+  uint64_t i = 0;
+  while (row.size() != 0) {
+    i++;
+    auto image = row["image"];
+    MS_LOG(INFO) << "Tensor image shape: " << image->shape();
+    iter->GetNextRow(&row);
+  }
+
+  EXPECT_EQ(i, 10);
+
+  // Manually terminate the pipeline
+  iter->Stop();
+}
+
+TEST_F(MindDataTestPipeline, TestRandomSolarizeSucess2) {
+  MS_LOG(INFO) << "Doing MindDataTestPipeline-TestRandomSolarize with default params.";
+  // Create an ImageFolder Dataset
+  std::string folder_path = datasets_root_path_ + "/testPK/data/";
+  std::shared_ptr<Dataset> ds = ImageFolder(folder_path, true, RandomSampler(false, 10));
+  EXPECT_NE(ds, nullptr);
+
+  // Create objects for the tensor ops
+  std::shared_ptr<TensorOperation> random_solarize = mindspore::dataset::api::vision::RandomSolarize();
+  EXPECT_NE(random_solarize, nullptr);
+
+  // Create a Map operation on ds
+  ds = ds->Map({random_solarize});
   EXPECT_NE(ds, nullptr);
 
   // Create an iterator over the result of the above dataset
@@ -1151,8 +1181,23 @@ TEST_F(MindDataTestPipeline, TestRandomSolarize) {
     iter->GetNextRow(&row);
   }
 
-  EXPECT_EQ(i, 20);
+  EXPECT_EQ(i, 10);
 
   // Manually terminate the pipeline
   iter->Stop();
 }
+
+TEST_F(MindDataTestPipeline, TestRandomSolarizeFail) {
+  MS_LOG(INFO) << "Doing MindDataTestPipeline-TestRandomSolarize with invalid params.";
+  std::vector<uint8_t> threshold = {13, 1};
+  std::shared_ptr<TensorOperation> random_solarize = mindspore::dataset::api::vision::RandomSolarize(threshold);
+  EXPECT_EQ(random_solarize, nullptr);
+
+  threshold = {1, 2, 3};
+  random_solarize = mindspore::dataset::api::vision::RandomSolarize(threshold);
+  EXPECT_EQ(random_solarize, nullptr);
+
+  threshold = {1};
+  random_solarize = mindspore::dataset::api::vision::RandomSolarize(threshold);
+  EXPECT_EQ(random_solarize, nullptr);
+}

tests/ut/cpp/dataset/random_solarize_op_test.cc

@@ -38,7 +38,9 @@ TEST_F(MindDataTestRandomSolarizeOp, TestOp1) {
   uint32_t curr_seed = GlobalContext::config_manager()->seed();
   GlobalContext::config_manager()->set_seed(0);
 
-  std::unique_ptr<RandomSolarizeOp> op(new RandomSolarizeOp(100, 100));
+  std::vector<uint8_t> threshold = {100, 100};
+  std::unique_ptr<RandomSolarizeOp> op(new RandomSolarizeOp(threshold));
+
   EXPECT_TRUE(op->OneToOne());
   Status s = op->Compute(input_tensor_, &output_tensor_);
 

tests/ut/cpp/dataset/solarize_op_test.cc

@@ -74,7 +74,8 @@ TEST_F(MindDataTestSolarizeOp, TestOp3) {
   MS_LOG(INFO) << "Doing testSolarizeOp3 - Pass in only threshold_min parameter";
 
   //  unsigned int threshold = 128;
-  std::unique_ptr<SolarizeOp> op(new SolarizeOp(1));
+  std::vector<uint8_t> threshold ={1, 255};
+  std::unique_ptr<SolarizeOp> op(new SolarizeOp(threshold));
 
   std::vector<uint8_t> test_vector = {3, 4, 59, 210, 255};
   std::vector<uint8_t> expected_output_vector = {252, 251, 196, 45, 0};
@@ -98,8 +99,8 @@ TEST_F(MindDataTestSolarizeOp, TestOp3) {
 TEST_F(MindDataTestSolarizeOp, TestOp4) {
   MS_LOG(INFO) << "Doing testSolarizeOp4 - Pass in both threshold parameters.";
 
-  //  unsigned int threshold = 128;
-  std::unique_ptr<SolarizeOp> op(new SolarizeOp(1, 230));
+  std::vector<uint8_t> threshold ={1, 230};
+  std::unique_ptr<SolarizeOp> op(new SolarizeOp(threshold));
 
   std::vector<uint8_t> test_vector = {3, 4, 59, 210, 255};
   std::vector<uint8_t> expected_output_vector = {252, 251, 196, 45, 255};
@@ -123,8 +124,8 @@ TEST_F(MindDataTestSolarizeOp, TestOp4) {
 TEST_F(MindDataTestSolarizeOp, TestOp5) {
   MS_LOG(INFO) << "Doing testSolarizeOp5 - Rank 2 input tensor.";
 
-  //  unsigned int threshold = 128;
-  std::unique_ptr<SolarizeOp> op(new SolarizeOp(1, 230));
+  std::vector<uint8_t> threshold ={1, 230};
+  std::unique_ptr<SolarizeOp> op(new SolarizeOp(threshold));
 
   std::vector<uint8_t> test_vector = {3, 4, 59, 210, 255};
   std::vector<uint8_t> expected_output_vector = {252, 251, 196, 45, 255};
@@ -149,7 +150,8 @@ TEST_F(MindDataTestSolarizeOp, TestOp5) {
 TEST_F(MindDataTestSolarizeOp, TestOp6) {
   MS_LOG(INFO) << "Doing testSolarizeOp6 - Bad Input.";
 
-  std::unique_ptr<SolarizeOp> op(new SolarizeOp(10, 1));
+  std::vector<uint8_t> threshold ={10, 1};
+  std::unique_ptr<SolarizeOp> op(new SolarizeOp(threshold));
 
   std::vector<uint8_t> test_vector = {3, 4, 59, 210, 255};
   std::shared_ptr<Tensor> test_input_tensor;

tests/ut/python/dataset/test_random_solarize_op.py

@@ -17,66 +17,87 @@ Testing RandomSolarizeOp op in DE
 """
 import pytest
 import mindspore.dataset as ds
+import mindspore.dataset.engine as de
 import mindspore.dataset.transforms.vision.c_transforms as vision
 from mindspore import log as logger
-from util import visualize_list, save_and_check_md5, config_get_set_seed, config_get_set_num_parallel_workers
+from util import visualize_list, save_and_check_md5, config_get_set_seed, config_get_set_num_parallel_workers, \
+    visualize_one_channel_dataset
 
 GENERATE_GOLDEN = False
 
+MNIST_DATA_DIR = "../data/dataset/testMnistData"
 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_random_solarize_op(threshold=None, plot=False):
+def test_random_solarize_op(threshold=(10, 150), plot=False, run_golden=True):
     """
     Test RandomSolarize
     """
     logger.info("Test RandomSolarize")
 
     # First dataset
-    data1 = ds.TFRecordDataset(DATA_DIR, SCHEMA_DIR, columns_list=["image"])
+    data1 = ds.TFRecordDataset(DATA_DIR, SCHEMA_DIR, columns_list=["image"], shuffle=False)
     decode_op = vision.Decode()
 
+    original_seed = config_get_set_seed(0)
+    original_num_parallel_workers = config_get_set_num_parallel_workers(1)
+
     if threshold is None:
         solarize_op = vision.RandomSolarize()
     else:
         solarize_op = vision.RandomSolarize(threshold)
+
     data1 = data1.map(input_columns=["image"], operations=decode_op)
     data1 = data1.map(input_columns=["image"], operations=solarize_op)
 
     # Second dataset
-    data2 = ds.TFRecordDataset(DATA_DIR, SCHEMA_DIR, columns_list=["image"])
+    data2 = ds.TFRecordDataset(DATA_DIR, SCHEMA_DIR, columns_list=["image"], shuffle=False)
     data2 = data2.map(input_columns=["image"], operations=decode_op)
 
+    if run_golden:
+        filename = "random_solarize_01_result.npz"
+        save_and_check_md5(data1, filename, generate_golden=GENERATE_GOLDEN)
+
     image_solarized = []
     image = []
+
     for item1, item2 in zip(data1.create_dict_iterator(), data2.create_dict_iterator()):
         image_solarized.append(item1["image"].copy())
         image.append(item2["image"].copy())
     if plot:
         visualize_list(image, image_solarized)
 
+    ds.config.set_seed(original_seed)
+    ds.config.set_num_parallel_workers(original_num_parallel_workers)
+
 
-def test_random_solarize_md5():
+def test_random_solarize_mnist(plot=False, run_golden=True):
     """
-    Test RandomSolarize
+    Test RandomSolarize op with MNIST dataset (Grayscale images)
     """
-    logger.info("Test RandomSolarize")
-    original_seed = config_get_set_seed(0)
-    original_num_parallel_workers = config_get_set_num_parallel_workers(1)
 
-    data1 = ds.TFRecordDataset(DATA_DIR, SCHEMA_DIR, columns_list=["image"], shuffle=False)
-    decode_op = vision.Decode()
-    random_solarize_op = vision.RandomSolarize((10, 150))
-    data1 = data1.map(input_columns=["image"], operations=decode_op)
-    data1 = data1.map(input_columns=["image"], operations=random_solarize_op)
-    # Compare with expected md5 from images
-    filename = "random_solarize_01_result.npz"
-    save_and_check_md5(data1, filename, generate_golden=GENERATE_GOLDEN)
+    mnist_1 = de.MnistDataset(dataset_dir=MNIST_DATA_DIR, num_samples=2, shuffle=False)
+    mnist_2 = de.MnistDataset(dataset_dir=MNIST_DATA_DIR, num_samples=2, shuffle=False)
+    mnist_2 = mnist_2.map(input_columns="image", operations=vision.RandomSolarize((0, 255)))
 
-    # Restore config setting
-    ds.config.set_seed(original_seed)
-    ds.config.set_num_parallel_workers(original_num_parallel_workers)
+    images = []
+    images_trans = []
+    labels = []
+
+    for _, (data_orig, data_trans) in enumerate(zip(mnist_1, mnist_2)):
+        image_orig, label_orig = data_orig
+        image_trans, _ = data_trans
+        images.append(image_orig)
+        labels.append(label_orig)
+        images_trans.append(image_trans)
+
+    if plot:
+        visualize_one_channel_dataset(images, images_trans, labels)
+
+    if run_golden:
+        filename = "random_solarize_02_result.npz"
+        save_and_check_md5(mnist_2, filename, generate_golden=GENERATE_GOLDEN)
 
 
 def test_random_solarize_errors():
@@ -105,8 +126,8 @@ def test_random_solarize_errors():
 
 
 if __name__ == "__main__":
-    test_random_solarize_op((100, 100), plot=True)
-    test_random_solarize_op((12, 120), plot=True)
-    test_random_solarize_op(plot=True)
+    test_random_solarize_op((10, 150), plot=True, run_golden=True)
+    test_random_solarize_op((12, 120), plot=True, run_golden=False)
+    test_random_solarize_op(plot=True, run_golden=False)
+    test_random_solarize_mnist(plot=True, run_golden=True)
     test_random_solarize_errors()
-    test_random_solarize_md5()