US 2026/0087786 A1 · SYSTEM, METHOD, AND PROGRAM PRODUCT FOR OUT OF DISTRIBUTION GENERALIZATION VIA INTERVENTIONAL STYLE TRANSFER

Abstract

The present disclosure relates to a method for generating a training set based on a first set of images and a second set of images, each image having a respective observational environment and a respective feature set. The method includes (a) obtaining, by a generator module, the first set of images and the second set of images; (b) extracting, by the generator module, one or more feature sets from each image in the first set of images; (c) extracting, by the generator module, one or more respective observational environments from each image in the second set of images; (d) deriving, by an encoder module, one or more latent representations from the one or more feature sets extracted from one or more images in the first set of images; (e) deriving, by the encoder module, one or more style codes from the one or more respective observational environment extracted from one or more images in the second set of images; (f) generating, by the generator module, an interventional training distribution having samples including each respective one or more style codes and each one or more latent representations; and, (g) storing, by the generator module, the interventional training distribution training set.

Claims20

Dependent claims

A method for generating a training set for a machine leaning predictor comprising:

a) obtaining, by a machine learning module, a content image and a style image;
b) extracting, by an encoder module, one or more features from the content image;
c) extracting, by the encoder module, an observational environment from the style image;
d) deriving, by the encoder module, one or more latent representations from the one or more feature sets extracted from the content image;
e) deriving, by the encoder module, at least one style code based on the observational environment extracted from the style image;
f) generating, by a generator module, an interventional training distribution training set element based on the style code and the latent representation;
g) storing, by the generator module, the interventional training distribution training set element in a memory operable connected to the generating module and the encoder module;
and
h) repeating steps a) through g) for a plurality of content images and a plurality of style images to provide an interventional training distribution training set.

6 more elements

An interventional style transfer system comprising:

a) at least one processor;
and
b) a memory, operably connected to the at least one processor, the memory includes processor executable code that when executed by the at least one processor executes steps of:
(i) obtaining a content image and a style image;
(ii) extracting one or more features from the content image;
(iii) extracting an observational environment from the style image;
(iv) deriving one or more latent representations from the one or more feature sets extracted from the content image;
(v) deriving at least one style code based on the observational environment extracted from the style image;
(vi) generating an interventional training distribution training set element based on the style code and the latent representation;
(vii) storing the interventional training distribution training set element in a memory operable connected to the generating module and the encoder module;
(viii) repeating steps (i) through (vii) for a plurality of content images and a plurality of style images to provide an interventional training distribution training set,
or
an interventional style transfer system comprising:
memory;
a machine learning module operatively connected to the memory and configured to obtain a content image and a style image;
an encoder module operably connected to the machine earning module and configured to:
(i) extract at least one feature set from the content image;
(ii) extract an observational environment from the style image;
(iii) derive a latent image representation based on the at least one feature set;
(iv) derive a style code based on the observational environment;
and
a generator module operatively connected to the encoder module and configured to generate an interventional training distribution training set element based on the latent image representation and the style code,
wherein the interventional training distribution training set element is stored in the memory to provide an interventional training distribution training set,
wherein the interventional training distribution training set is used to train a machine learning predictor.

21 more elements

Dependencies are inferred from the claim text.

System, method, and program product for out of distribution generalization via interventional style transfer

Life of this patent

System, method, and program product for out of distribution generalization via interventional style transfer

Life of this patent

Abstract

Claims20