×

Generative methods of super resolution

  • US 10,692,185 B2
  • Filed: 12/28/2017
  • Issued: 06/23/2020
  • Est. Priority Date: 03/18/2016
  • Status: Active Grant
First Claim
Patent Images

1. A method for training an algorithm to process at least a section of received visual data using a training dataset and a reference dataset, the method being an iterative method with each iteration comprising:

  • generating a set of training data from the training dataset using the algorithm;

    determining one or more characteristics of the training data, wherein the one or more characteristics include a statistical distribution of the training data;

    determining one or more characteristics of the reference dataset, wherein the one or more characteristics include a statistical distribution of the reference dataset;

    comparing the one or more characteristics of the training data to the one or more characteristics of the reference dataset; and

    modifying one or more parameters of the algorithm to optimise processed visual data based on the comparison between the one or more characteristics of the training data and the one or more characteristics of the reference dataset,wherein the algorithm outputs the processed visual data with the same content as the at least a section of received visual data.

View all claims
  • 1 Assignment
Timeline View
Assignment View
    ×
    ×