Generative augmentation of image data

Abstract

Systems and methods to receive one or more first images associated with a training set of images to train a machine learning model; provide the one or more first images as a first input to a first set of layers of computational units, wherein the first set of layers utilizes image filters; provide a first output of the first set of layers of computational units as a second input to a second layer of the computational units, wherein the second layer utilizes random parameter sets for computations; obtain distortion parameters from the second layer of the computational units; generate one or more second images comprising a representation of the one or more first images modified with the distortion parameters; obtain, as a third output, the one or more second images; and add the one or more second images to the training set of images to train the machine learning model.

What is claimed is: 
     
         1 . A method comprising:
 providing a representation of a first image associated with a machine learning model to a first layer of a set of computational units, wherein the first layer utilizes random parameter sets for computations;   obtaining distortion parameters from the first layer of the set of computational units; and   generating a second image based on the first image and the distortion parameters, the second image generated to be used in the machine learning model.   
     
     
         2 . The method of  claim 1 , further comprising:
 providing an output of the first layer of the set of computational units as an input to a second layer of the set of computational units.   
     
     
         3 . The method of  claim 1 , wherein the machine learning model comprises a convolutional neural network. 
     
     
         4 . The method of  claim 1 , wherein the representation of the first image is obtained by:
 receiving the first image, wherein the first image is associated with a training set of images to train the machine learning model;   dividing the first image into a plurality of portions;   providing each of the plurality of portions to a third layer of the set of computational units; and   obtaining an output of the third layer as the representation of the first image.   
     
     
         5 . The method of  claim 4 , wherein obtaining the distortion parameters comprises:
 obtaining a distortion parameter for each of the plurality of portions of the first image.   
     
     
         6 . The method of  claim 1 , further comprising:
 generating a randomized kernel matrix for the first layer of the set of computational units based on one or more matrices.   
     
     
         7 . The method of  claim 6 , wherein the one or more matrices comprise one or more of:
 a first matrix of mean values, the first matrix initialized with random values;   a second matrix of standard deviation values, the second matrix initialized with zero values;   a third matrix of displacement values, the third matrix based on a number of filters to apply to the representation of the first image; or   a fourth matrix that is based on an arbitrary standard deviation value and a normal distribution value.   
     
     
         8 . The method of  claim 7 , wherein the arbitrary standard deviation value specifies a roughness of image distortions. 
     
     
         9 . The method of  claim 6 , wherein parameters for the one or more matrices comprise at least one of filter height, filter width, image height, image width, size of filter, number of channels, number of filters, or number of images. 
     
     
         10 . The method of  claim 9 , wherein the image height and image width each comprises arbitrary values. 
     
     
         11 . The method of  claim 1 , wherein the first image comprises one or more of:
 one or more hieroglyphs;   one or more Chinese-Japanese-Korean (CJK) symbols;   one or more Arabic strings; or   a combination of one or more other symbols.   
     
     
         12 . The method of  claim 1 , wherein generating the second image comprises:
 generating the second image corresponding to naturally distorted images.   
     
     
         13 . The method of  claim 1 , further comprising: adding the second image to a training set of images to train the machine learning model. 
     
     
         14 . The method of  claim 7 , wherein the first matrix, the second matrix, and the third matrix each comprises learnable parameters. 
     
     
         15 . A system comprising:
 a memory; and   a processor, coupled to the memory, the processor to:
 provide a representation of a first image associated with a machine learning model to a first layer of a set of computational units, wherein the first layer utilizes random parameter sets for computations; 
 obtain distortion parameters from the first layer of the set of computational units; and 
 generate a second image based on the first image and the distortion parameters, the second image generated to be used in the machine learning model 
   
     
     
         16 . The system of  claim 15 , wherein the processor is further to:
 generate a randomized kernel matrix for the first layer of the set of computational units based on one or more matrices.   
     
     
         17 . The system of  claim 16 , wherein the one or more matrices comprise one or more of:
 a first matrix of mean values, the first matrix initialized with random values;   a second matrix of standard deviation values, the second matrix initialized with zero values;   a third matrix of displacement values, the third matrix based on a number of filters to apply to the representation of the first image; or   a fourth matrix that is based on an arbitrary standard deviation value and a normal distribution value.   
     
     
         18 . The system of  claim 17 , wherein the arbitrary standard deviation value specifies roughness of image distortions. 
     
     
         19 . A computer-readable non-transitory storage medium comprising executable instructions that, when executed by a processing device, cause the processing device to:
 provide a representation of a first image associated with a machine learning model to a first layer of a set of computational units, wherein the first layer utilizes random parameter sets for computations;   obtain distortion parameters from the first layer of the set of computational units; and   generate a second image comprising the representation of the first image modified with the distortion parameters, the second image generated to be used in the machine learning model.   
     
     
         20 . The computer-readable non-transitory storage medium of  claim 19 , wherein the processing device is further to:
 generate a randomized kernel matrix for the first layer of the set of computational units based on one or more matrices.