Class AutoAugment

Inheritance Relationships

Base Type

Class Documentation

class AutoAugment : public mindspore::dataset::TensorTransform

Apply AutoAugment data augmentation method.

Public Functions

explicit AutoAugment(AutoAugmentPolicy policy = AutoAugmentPolicy::kImageNet, InterpolationMode interpolation = InterpolationMode::kNearestNeighbour, const std::vector<uint8_t> &fill_value = {0, 0, 0})

Constructor.

Example
/* Define operations */
auto decode_op = vision::Decode();
auto auto_augment_op = vision::AutoAugment(AutoAugmentPolicy::kImageNet,
                                           InterpolationMode::kNearestNeighbour, {0, 0, 0});
/* dataset is an instance of Dataset object */
dataset = dataset->Map({decode_op, auto_augment_op}, // operations
                       {"image"});                   // input columns

Parameters

  • policy – [in] An enum for the data auto augmentation policy (default=AutoAugmentPolicy::kImageNet).

    • AutoAugmentPolicy::kImageNet, AutoAugment policy learned on the ImageNet dataset.

    • AutoAugmentPolicy::kCifar10, AutoAugment policy learned on the Cifar10 dataset.

    • AutoAugmentPolicy::kSVHN, AutoAugment policy learned on the SVHN dataset.

  • interpolation – [in] An enum for the mode of interpolation (default=InterpolationMode::kNearestNeighbour).

    • InterpolationMode::kNearestNeighbour, Interpolation method is nearest-neighbor interpolation.

    • InterpolationMode::kLinear, Interpolation method is blinear interpolation.

    • InterpolationMode::kCubic, Interpolation method is bicubic interpolation.

    • InterpolationMode::kArea, Interpolation method is pixel area interpolation.

  • fill_value – [in] A vector representing the pixel intensity of the borders (default={0, 0, 0}).

~AutoAugment() = default

Destructor.