Automated design testing through deep learning

1. A method of evaluating a test subject, the method comprising:

generating, by a first machine learning system, test subject features;

analyzing, by a second machine learning system, the test subject to detect distinguishing features of the test subject;

performing, by a third machine learning system, natural language processing on the test subject features to create evaluation information associated with the test subject; and

providing an evaluation of the test subject based on the distinguishing features and the evaluation information.

2. The method of claim 1 , wherein the first machine learning system is a Convolutional Neural Network, and wherein the Convolutional Neural Network is trained using a plurality of images, wherein at least one of the plurality of images is translated into multiple representations of the at least one of the plurality of images, wherein the Convolutional Neural Network is trained with the multiple representations.

3. The method of claim 1 , wherein generating, by the first machine learning system, test subject features comprises:

receiving a plurality of images of the test subject, wherein the plurality of images comprise a plurality of representations of the test subject;

creating a three-dimensional view of the plurality of images; and

receiving, by the first machine learning system, the three-dimensional view as input.

4. The method of claim 3 , wherein the input to the second machine learning system is the three-dimensional view.

5. The method of claim 1 , wherein the second machine learning system is a Generative Adversarial Network, and wherein the Generative Adversarial Network is trained using training images comprising normal and abnormal test subjects.

6. The method of claim 1 , wherein the detected distinguishing features are anomalies present in the test subject.

7. The method of claim 1 , wherein the second machine learning system outputs a result for the test subject of normal or abnormal.

8. The method of claim 1 , wherein the input to the third machine learning system is the output of the first machine learning system.

9. The method of claim 1 , wherein the third machine learning system is a Recurrent Neural Network, and wherein the Recurrent Neural Network is trained to interpret vector information provided as output by the first machine learning system.

10. A system for use in evaluating a test subject in computing environments, the system comprising a processor configured to:

generate, by a first machine learning system, test subject features;

analyze, by a second machine learning system, the test subject to detect distinguishing features of the test subject;

perform, by a third machine learning system, natural language processing on the test subject features to create evaluation information associated with the test subject; and

provide an evaluation of the test subject based on the distinguishing features and the evaluation information.

11. The system of claim 10 , wherein the first machine learning system is a Convolutional Neural Network, and wherein the Convolutional Neural Network is trained using a plurality of images, wherein at least one of the plurality of images is translated into multiple representations of the at least one of the plurality of images, wherein the Convolutional Neural Network is trained with the multiple representations.

12. The system of claim 10 , wherein the processor configured to generate, by the first machine learning system, test subject features, is further configured to:

receive a plurality of images of the test subject, wherein the plurality of images comprise a plurality of representations of the test subject;

create a three-dimensional view of the plurality of images; and

receive, by the first machine learning system, the three-dimensional view as input.

13. The system of claim 12 , wherein the input to the second machine learning system is the three-dimensional view.

14. The system of claim 10 , wherein the second machine learning system is a Generative Adversarial Network, and wherein the Generative Adversarial Network is trained using training images comprising normal and abnormal test subjects.

15. The system of claim 10 , wherein the detected distinguishing features are anomalies present in the test subject.

16. The system of claim 10 , wherein the second machine learning system outputs a result for the test subject of normal or abnormal.

17. The system of claim 10 , wherein the input to the third machine learning system is the output of the first machine learning system.

18. The system of claim 10 , wherein the third machine learning system is a Recurrent Neural Network, and wherein the Recurrent Neural Network is trained to interpret vector information provided as output by the first machine learning system.

19. A computer program product for evaluating a test subject in computing environments, the computer program product comprising:

a non-transitory computer readable storage medium having computer executable program code embodied therewith, the program code executable by a computer processor to:

generate, by a first machine learning system, test subject features;

analyze, by a second machine learning system, the test subject to detect distinguishing features of the test subject;

perform, by a third machine learning system, natural language processing on the test subject features to create evaluation information associated with the test subject; and

provide an evaluation of the test subject based on the distinguishing features and the evaluation information.

20. The computer program product of claim 19 , wherein the first machine learning system is a Convolutional Neural Network, and wherein the Convolutional Neural Network is trained using a plurality of images, wherein at least one of the plurality of images is translated into multiple representations of the at least one of the plurality of images, wherein the Convolutional Neural Network is trained with the multiple representations.