Method and arrangement for image model construction

A method for constructing an image model from at least one image data input, the method comprising the steps of:

• determining at least one state of the at least one image data input, and a state of an intermediate learning model;
• determining a target state from the at least one state of the at least one image data input, and from the state of the intermediate learning model;
• performing at least one transformation on the at least one image data input using the determined target state to generate at least one transformed image;
• aggregating the at least one transformed image with intermediate learning model information to generate an updated estimate of the image model;
• providing the updated estimate of the image model as the image model;
• and
• using the updated estimate of the image model to generate an update of the intermediate learning model.

An apparatus configured to perform a method comprising the steps of:

• determining at least one state of the at least one image data input, and a state of an intermediate learning model;
• determining a target state from the at least one state of the at least one image data input, and from the state of the intermediate learning model;
• performing at least one transformation on the at least one image data input using the determined target state to generate at least one transformed image;
• aggregating the at least one transformed image with intermediate learning model information to generate an updated estimate of the image model;
• providing the updated estimate of the image model as the image model;
• and
• using the updated estimate of the image model to generate an update of the intermediate learning model.

A digital data storage medium encoding a machine-executable program of instructions to perform a method, the method comprising the steps of:

• determining at least one state of the at least one image data input, and a state of an intermediate learning model;
• determining a target state from the at least one state of the at least one image data input, and from the state of the intermediate learning model;
• performing at least one transformation on the at least one image data input using the determined target state to generate at least one transformed image;
• aggregating the at least one transformed image with intermediate learning model information to generate an updated estimate of the image model;
• providing the updated estimate of the image model as the image model;
• and using the updated estimate of the image model to generate an update of the intermediate learning model.