Function mindspore::dataset::Cityscapes

Function Documentation

inline std::shared_ptr<CityscapesDataset> mindspore::dataset::Cityscapes(const std::string &dataset_dir, const std::string &usage, const std::string &quality_mode, const std::string &task, bool decode = false, const std::shared_ptr<Sampler> &sampler = std::make_shared<RandomSampler>(), const std::shared_ptr<DatasetCache> &cache = nullptr)

Function to create a CityscapesDataset.

Note

The generated dataset has two columns [“image”, “task”].

Parameters
  • dataset_dir[in] The dataset dir to be read.

  • usage[in] The type of dataset. Acceptable usages include “train”, “test”, “val” or “all” if quality_mode is “fine” otherwise “train”, “train_extra”, “val” or “all”.

  • quality_mode[in] The quality mode of processed image. Acceptable quality_modes include “fine” or “coarse”.

  • task[in] The type of task which is used to select output data. Acceptable tasks include “instance”, “semantic”, “polygon” or “color”.

  • decode[in] Decode the images after reading (default=false).

  • sampler[in] Shared pointer to a sampler object used to choose samples from the dataset. If sampler is not given, a RandomSampler will be used to randomly iterate the entire dataset (default = RandomSampler()).

  • cache[in] Tensor cache to use. (default=nullptr which means no cache is used).

Returns

Shared pointer to the current CityscapesDataset.

Example
/* Define dataset path and MindData object */
std::string folder_path = "/path/to/cityscapes_dataset_directory";
std::shared_ptr<Dataset> ds = Cityscapes(dataset_path, "train", "fine", "color");

/* Create iterator to read dataset */
std::shared_ptr<Iterator> iter = ds->CreateIterator();
std::unordered_map<std::string, mindspore::MSTensor> row;
iter->GetNextRow(&row);

/* Note: In Cityscapes dataset, each data dictionary owns keys "image" and "task" */
auto task = row["task"];