Type Members

1.  class BytesToMat extends FeatureTransformer

   Transform byte array(original image file in byte) to OpenCVMat

2.  class ChainedFeatureTransformer extends FeatureTransformer

   A transformer chain two FeatureTransformer together.

3.  class ClassificationMTImageFeatureToBatch extends MTImageFeatureToBatch

   A transformer pipeline wrapper to create labeled Minibatch in multiple threads for classification

4.  class DistributedImageFrame extends ImageFrame

   Distributerd ImageFrame, it keeps an rdd of ImageFeature

5.  abstract class FeatureTransformer extends Transformer[ImageFeature, ImageFeature]

   FeatureTransformer is a transformer that transform ImageFeature

6.  class ImageFeature extends Serializable

   Each ImageFeature keeps information about single image, it can include various status of an image, e.g.

   Each ImageFeature keeps information about single image, it can include various status of an image, e.g. original bytes read from image file, an opencv mat, pixels in float array, image label, meta data and so on. it uses HashMap to store all these data, the key is string that identify the corresponding value

7.  class ImageFeatureToMiniBatch[T] extends Transformer[ImageFeature, MiniBatch[T]]
8.  trait ImageFrame extends Serializable

   ImageFrame wraps a set of ImageFeature

9.  class ImageFrameToSample[T] extends FeatureTransformer

   Transforms tensors that map inputKeys and targetKeys to sample

10.  class LocalImageFrame extends ImageFrame

    Local ImageFrame, keeps an array of ImageFeature

11.  abstract class MTImageFeatureToBatch extends Transformer[ImageFeature, MiniBatch[Float]]

    An abstract class to convert ImageFeature iterator to MiniBatches.

    An abstract class to convert ImageFeature iterator to MiniBatches. This transformer will be run on each image feature. "processImageFeature" will be called to buffer the image features. When there are enough buffered image features to form a batch, "createBatch" will be called. You should override processImageFeature to buffer each image feature, and createBatch to convert the buffered data into a mini-batch

12.  class MatToFloats extends FeatureTransformer

    Transform OpenCVMat to float array, note that in this transformer, the mat is released

13.  class MatToTensor[T] extends FeatureTransformer

    transform opencv mat to tensor

14.  class PixelBytesToMat extends FeatureTransformer

    Transform byte array(pixels in byte) to OpenCVMat

15.  class RoiImageFeatureToBatch extends MTImageFeatureToBatch

    A transformer pipeline wrapper to create RoiMiniBatch in multiple threads The output "target" is a Table.

    A transformer pipeline wrapper to create RoiMiniBatch in multiple threads The output "target" is a Table. The keys are from 1 to sizeof(batch). The values are the tables for each RoiLabel. Each Roi label table, contains fields of RoiLabel class. The sizes of the input images should be the same

16.  class RoiImageFeatureToBatchWithResize extends MTImageFeatureToBatch

    A transformer pipeline wrapper to create RoiMiniBatch in multiple threads.

    A transformer pipeline wrapper to create RoiMiniBatch in multiple threads. Image features may have different sizes, so firstly we need to calculate max size in one batch, then padding all features to one batch with max size.

17.  class RoiMiniBatch extends MiniBatch[Float]

    A batch of images with flattened RoiLabels the getTarget() returns a Table with key from 1 to batchSize.

    A batch of images with flattened RoiLabels the getTarget() returns a Table with key from 1 to batchSize. Each key in the table is mapped to a Table for the annotation of an image in the batch. The annotation table holds the annotation info for one image (assume the image has N detections). The annotation table has

    Key Value RoiImageInfo.CLASSES the categories for each detections (see RoiLabel.clasees field) (1 x N), or (2 x N) Tensor[Float] RoiImageInfo.BBOXES the bboxes, (N x 4) Tensor[Float] RoiImageInfo.MASKS (Optional) the mask data, Array[Tensor[Float]\]. The outer array has N elements. The inner tensor holds the data for segmentation RoiImageInfo.ISCROWD Whether each detection is crowd. (1 x N) Tensor[Float]. -1: unknown, 0: not crowd, 1: is crowd RoiImageInfo.IMGINFO with shape (batchSize, 4), contains all images info (height, width, original height, original width)