# US 2021/0279869 A1 · Deep neural network based identification of realistic synthetic images generated using a generative adversarial network

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2021/0279869 A1
- Document ID: US20210279869A1
- Status: Published application
- Kind: A1
- Application number: 17/327,239
- Filed: 2021-05-21
- Published: 2021-09-09
- Earliest claimed: 2019-03-29
- Assignees: GE Precision Healthcare LLC
- Applicants: GE Precision Healthcare LLC
- Inventors: Ravi Soni; Min Zhang; Zili Ma; Gopal B. Avinash
- Web page: https://patentgrep.com/patent/US20210279869A1

## CPC

- G16H 50/20: for computer-aided diagnosis, e.g. based on medical expert systems
- G06N 3/045: Combinations of networks
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/047: Probabilistic or stochastic networks
- G06N 3/0475: Generative networks
- G06N 3/08: Learning methods
- G06N 3/09: Supervised learning
- G06N 3/094: Adversarial learning
- G06T 7/0012: {Biomedical image inspection}
- G06V 10/772: Determining representative reference patterns, e.g. averaging or distorting patterns; Generating dictionaries
- G06V 10/774: Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
- G06V 10/7784: {based on feedback from supervisors}
- G06V 10/82: using neural networks
- G16H 30/40: for processing medical images, e.g. editing
- G06T 2207/20081: Training; Learning
- G06T 2207/20084: Artificial neural networks [ANN]
- G06T 2207/20092: Interactive image processing based on input by user
- G06V 2201/03: Recognition of patterns in medical or anatomical images
- G16H: HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

Techniques are provided for deep neural network (DNN) identification of realistic synthetic images generated using a generative adversarial network (GAN). According to an embodiment, a system is described that can comprise a memory that stores computer executable components and a processor that executes the computer executable components stored in the memory. The computer executable components can comprise, a first extraction component that extracts a subset of synthetic images classified as non-real like as opposed to real-like, wherein the subset of synthetic images were generated using a GAN model. The computer executable components can further comprise a training component that employs the subset of synthetic images and real images to train a DNN network model to classify synthetic images generated using the GAN model as either real-like or non-real like

## Claims

### Claim 1

A system, comprising:
a memory that stores computer executable components; and
a processor that executes the computer executable components stored in the memory, wherein the computer executable components comprise:
an inference component that employs a deep neural network model to classify synthetic images as either real-like or non-real-like; and
a target model training component that employs a subset of the synthetic images classified as real-like to train a machine learning model to perform artificial intelligence analytics on real images.

### Claim 2 (depends on 1)

The system of claim 1, wherein the real images and the synthetic images comprise medical images.

### Claim 3 (depends on 1)

The system of claim 1, wherein the synthetic images were generated using a generative adversarial network model.

### Claim 4 (depends on 1)

The system of claim 1, wherein the computer executable components further comprise:
a synthetic image generator that generates the synthetic images using a generative adversarial network model.

### Claim 5 (depends on 1)

The system of claim 1, wherein the synthetic images comprise second synthetic images and the real images comprise second real images, and wherein the deep neural network model was trained on a training dataset comprising first real images and first synthetic images classified as non-real like.

### Claim 6 (depends on 5)

The system of claim 5, wherein the first synthetic images and the second synthetic images were generated using a generative adversarial network model.

### Claim 7 (depends on 5)

The system of claim 5, wherein the computer executable components further comprise:
a training component that employs the training dataset to train the deep neural network model to classify test synthetic images as either real-like or non-real like.

### Claim 8 (depends on 1)

The system of claim 1, wherein the deep neural network model comprises a very deep convolutional neural network model configured for binary classification.

### Claim 9 (depends on 2)

The system of claim 2, wherein the artificial intelligence analytics comprise automated evaluation of one or more medical conditions reflected in the real images.

### Claim 10

A method, comprising:
classifying, by a system operatively coupled to a processor, synthetic images as either real-like or non-real-like using a deep neural network model; and
employing, by the system, a subset of the synthetic images classified as real-like to train a machine learning model to perform artificial intelligence analytics on real images.

### Claim 11 (depends on 10)

The method of claim 10, wherein the real images and the synthetic images comprise medical images.

### Claim 12 (depends on 10)

The method of claim 10, wherein the synthetic images were generated using a generative adversarial network model.

### Claim 13 (depends on 10)

The method of claim 10, further comprising:
generating, by the system, the synthetic images using a generative adversarial network model.

### Claim 14 (depends on 10)

The method of claim 10, wherein the synthetic images comprise second synthetic images and the real images comprise second real images, and wherein the deep neural network model was trained on a training dataset comprising first real images and first synthetic images classified as non-real like.

### Claim 15 (depends on 14)

The method of claim 14, wherein the first synthetic images and the second synthetic images were generated using a generative adversarial network model.

### Claim 16 (depends on 14)

The method of claim 14, further comprising:
employing, by the system, the training dataset to train the deep neural network model to classify test synthetic images as either real-like or non-real like.

### Claim 17 (depends on 10)

The method of claim 10, wherein the deep neural network model comprises a very deep convolutional neural network model configured for binary classification.

### Claim 18 (depends on 11)

The method of claim 11, wherein the artificial intelligence analytics comprise automated evaluation of one or more medical conditions reflected in the real images.

### Claim 19

A machine-readable storage medium, comprising executable instructions that, when executed by a processor, facilitate performance of operations, comprising:
classifying synthetic images as either real-like or non-real-like using a deep neural network model; and
employing a subset of the synthetic images classified as real-like to train a machine learning model to perform artificial intelligence analytics on real images.

### Claim 20 (depends on 19)

The machine-readable storage medium of claim 19, wherein the synthetic images comprise second synthetic images and the real images comprise second real images, and wherein the deep neural network model was trained on a training dataset comprising first real images and first synthetic images classified as non-real like.