Enhanced resolution video and security via machine learning

Patent metadata

Field	Value
Publication number	US-10271008-B2
Application number	US-201715485071-A
Country	US
Kind code	B2
Filing date	Apr 11, 2017
Priority date	Apr 11, 2017
Publication date	Apr 23, 2019
Grant date	Apr 23, 2019

Abstract

Systems, apparatuses, and methods for enhanced resolution video and security via machine learning are disclosed. A transmitter reduces a resolution of each image of a videostream from a first, higher image resolution to a second, lower image resolution. The transmitter generates a set of parameters for programming a neural network to reconstruct a version of each image at the first image resolution. Then, the transmitter sends the images at the second image resolution to the receiver, along with the first set of parameters. The receiver programs a neural network with the first set of parameters and uses the neural network to reconstruct versions of the images at the first image resolution. The transmitter can send the first set of parameters to the receiver via a secure channel, ensuring that only the receiver can decode the images from the second image resolution to the first image resolution.

First claim

What is claimed is:

1. A system comprising:
a memory; and
a processor coupled to the memory;
wherein the system is configured to:
receive, via a secure channel, a first set of parameters for programming a neural network;
receive an image, via an insecure channel, wherein prior to being received, secure data has been encoded in the image and a resolution of the image has been reduced from a first image resolution to a second image resolution;
program the neural network with the first set of parameters; and
process the image of the second image resolution with the neural network to:
reconstruct a version of the image with the first image resolution; and
extract the secure data from the image with the first image resolution;
wherein said secure data cannot be recovered without said parameters.

2. The system as recited in claim 1 , wherein the image has been partitioned into a plurality of regions.

3. The system as recited in claim 2 , wherein the system is further configured to:
receive a plurality of sets of parameters, wherein a different set of parameters corresponds to each region of the plurality of regions;
program a plurality of neural networks with the plurality of sets of parameters, with each neural network programmed by a separate set of parameters corresponding to a different region of the plurality of regions; and
process each region of the plurality of regions with a separate neural network to reconstruct a version of the region of the first image resolution.

4. The system as recited in claim 3 , wherein the system is further configured to combine the plurality of regions to reconstruct the version of the image of the first image resolution.

5. The system as recited in claim 1 , wherein the secure data is obscured by reducing the image from the first image resolution to the second image resolution.

6. The system as recited in claim 5 , wherein the system is further configured to extract the secure data from the version of the image of the first image resolution.

7. The system as recited in claim 1 , wherein the image is part of a videostream.

8. A method comprising:
receive, via a secure channel, a first set of parameters for programming a neural network;
receive an image, via an insecure channel, wherein prior to being received, secure data has been encoded in the image and a resolution of the image has been reduced from a first image resolution to a second image resolution;
programming the neural network with the first set of parameters; and
processing the image of the second image resolution with the neural network to:
reconstruct a version of the image with the first image resolution; and
extract the secure data from the image with the first image resolution;

wherein said secure data cannot be recovered without said parameters.

9. The method as recited in claim 8 , wherein the image has been partitioned into a plurality of regions.

10. The method as recited in claim 9 , further comprising:
receiving a plurality of sets of parameters, wherein a different set of parameters corresponds to each region of the plurality of regions;
programming a plurality of neural networks with the plurality of sets of parameters, with each neural network programmed by a separate set of parameters corresponding to a different region of the plurality of regions; and
processing each region of the plurality of regions with a separate neural network to reconstruct a version of the region of the first image resolution.

11. The method as recited in claim 10 , further comprising combining the plurality of regions to reconstruct the version of the image of the first image resolution.

12. The method as recited in claim 8 , wherein the secure data is obscured by reducing the image from the first image resolution to the second image resolution.

13. The method as recited in claim 12 , further comprising extracting the secure data from the version of the image of the first image resolution.

14. The method as recited in claim 8 , wherein the image is part of a videostream.

15. An apparatus comprising:
a memory;
a processor; and
a neural network;
wherein the apparatus is configured to:
receive, via a secure channel, a first set of parameters for programming the neural network;
receive an image, via an insecure channel, wherein prior to being received, secure data has been encoded in the image and a resolution of the image has been reduced from a first image resolution to a second image resolution;
program the neural network with the first set of parameters; and
process the image of the second image resolution with the neural network to:
reconstruct a version of the image with the first image resolution; and
extract the secure data from the image with the first image resolution;
wherein said secure data cannot be recovered without said parameters.

16. The apparatus as recited in claim 15 , wherein the image has been partitioned into a plurality of regions.

17. The apparatus as recited in claim 16 , wherein the apparatus is further configured to:
receive a plurality of sets of parameters, wherein a different set of parameters corresponds to each region of the plurality of regions;
program a plurality of neural networks with the plurality of sets of parameters, with each neural network programmed by a separate set of parameters corresponding to a different region of the plurality of regions; and
process each region of the plurality of regions with a separate neural network to reconstruct a version of the region of the first image resolution.

18. The apparatus as recited in claim 17 , wherein the apparatus is further configured to combine the plurality of regions to reconstruct the version of the image of the first image resolution.

19. The apparatus as recited in claim 15 , wherein the secure data is obscured by reducing the image from the first image resolution to the second image resolution.

20. The apparatus as recited in claim 19 , wherein the apparatus is further configured to extract the secure data from the version of the image of the first image resolution.

Assignees

• Advanced Micro Devices Inc

Inventors

• Breternitz Mauricio
• Owen Jr John E

Classifications

• G06N3/045

  Combinations of networks · CPC title

• G06T3/40

  Scaling of whole images or parts thereof, e.g. expanding or contracting · CPC title

• H04L2209/34

  Encoding or coding, e.g. Huffman coding or error correction · CPC title

• H04N7/0117 Primary

  involving conversion of the spatial resolution of the incoming video signal (for graphics images G09G2340/0407) · CPC title

• H04L63/0428

  wherein the data content is protected, e.g. by encrypting or encapsulating the payload · CPC title