Style transfer

Patent metadata

Field	Value
Publication number	US-12293142-B2
Application number	US-202318106802-A
Country	US
Kind code	B2
Filing date	Feb 7, 2023
Priority date	Mar 4, 2019
Publication date	May 6, 2025
Grant date	May 6, 2025

Abstract

Various implementations of the present disclosure relate to style transfer. In some implementations, a computer-implemented method comprises: obtaining a target object having a first style, a style of the target object being editable; obtaining a reference image including a reference object; obtaining a second style of the reference object, the second style of the reference object being extracted from the reference image; and applying the second style to the target object.

First claim

The invention claimed is:

1. A system comprising:
at least one processor; and
memory comprising instructions that, when executed by the at least one processor, cause the at least one processor to perform operations to:
obtain a target editable object, from a presentation computer application, the target editable object associated with a first data set;
calculate a similarity measure for a plurality of predefined editable objects through comparison of the plurality of predefined editable objects to the target editable object, each predefined editable object of the plurality of predefined editable objects associated with a second data set, wherein the similarity is measured based on data of the first data set and the second data set, semantic information associated with the first data set and the second data set, or categories of the first data set and the second data set;
identify one or more predefined editable objects from the plurality of predefined editable objects based on a similarity measurement of the one or more predefined editable objects being outside a threshold similarity measure;
extract a style of the one or more predefined editable objects by a style parser of a neural network;
output the one or more predefined editable objects in a user interface of the presentation computer application; and
upon receipt of a selection of the one or more predefined editable objects, apply the style of the one or more predefined editable objects to the target editable object by a style adapter of the neural network in an application window of the presentation computer application, wherein the style of the one or more predefined editable objects comprises at least one visual display characteristic.

2. The system of claim 1 , the memory further comprising instructions that, when executed by the at least one processor, cause the at least one processor to perform operations to generate the target editable object from the data set.

3. The system of claim 1 , wherein the target editable object is an electronic chart that comprises a visualization of the data set.

4. The system of claim 1 , wherein a predefined editable object of the plurality of predefined editable objects has a respective style.

5. The system of claim 1 , the instructions to compare the plurality of predefined editable objects to the target editable object further comprising instructions that, when executed by the at least one processor, cause the at least one processor to perform operations to compare at least one of a data size, a row count, a column count, a data value, a category, or text content.

6. The system of claim 1 , the instructions to apply the style to the target editable object further comprising instructions that, when executed by the at least one processor, cause the at least one processor to perform operations to alter a current display characteristic of the target editable object to match the at least one visual display characteristic.

7. A computer-implemented method comprising:
obtaining a target editable object associated with a first data set, from a presentation computer application, the target editable object;
calculating a similarity measure for a plurality of predefined editable objects by comparing the plurality of predefined editable objects to the target editable object, each predefined editable object of the plurality of predefined editable objects associated with a second data set, wherein the similarity is measured based on data of the first data set and the second data set, semantic information associated with the first data set and the second data set, or categories of the first data set and the second data set;
identifying one or more predefined editable objects from the plurality of predefined editable objects based on a similarity measurement of the one or more predefined editable objects being outside a threshold similarity measure;
extracting a style of the one or more predefined editable objects by a style parser of a neural network;
outputting the one or more predefined editable objects in a user interface of the presentation computer application; and
upon receipt of a selection of the one or more predefined editable objects, applying the style of the one or more predefined editable objects to the target editable object by a style adapter of the neural network in an application window of the presentation computer application, wherein the style of the one or more predefined editable objects comprises at least one visual display characteristic.

8. The computer-implemented method of claim 7 , further comprising generating the target editable object from the data set.

9. The computer-implemented method of claim 7 , wherein the target editable object is an electronic chart that comprises a visualization of the data set.

10. The computer-implemented method of claim 7 , wherein a predefined editable object of the plurality of predefined editable objects has a respective style.

11. The computer-implemented method of claim 7 , wherein comparing the plurality of predefined editable objects to the target editable object includes comparing at least one of a data size, a row count, a column count, a data value, a category, or text content.

12. The computer-implemented method of claim 7 , wherein applying the style to the target editable object comprises altering a current display characteristic of the target editable object to match the at least one visual display characteristic.

13. At least one non-transitory machine-readable medium comprising instructions that, when executed by at least one processor, cause the at least one processor to perform operations to:
obtain a target editable object associated with a first data set, from a presentation computer application, the target editable object;
calculate a similarity measure for a plurality of predefined editable objects through comparison of the plurality of predefined editable objects to the target editable object, each predefined editable object of the plurality of predefined editable objects associated with a second data set, wherein the similarity is measured based on data of the first data set and the second data set, semantic information associated with the first data set and the second data set, or categories of the first data set and the second data set;
identify one or more predefined editable objects from the plurality of predefined editable objects based on a similarity measurement of the one or more predefined editable objects being outside a threshold similarity measure;
extracting a style of the one or more predefined editable objects by a style parser of a neural network;
output the one or more predefined editable objects in a user interface of the presentation computer application; and
upon receipt of a selection of the one or more predefined editable objects, apply the style of the one or more predefined editable objects to the target editable object by a style adapter of the neural network in an application window of the presentation computer application, wherein the style of the one or more predefined editable objects comprises at least one visual display characteristic.

14. The at least one non-transitory machine-readable medium of claim 13 , further comprising instructions that, when executed by the at least one processor, cause the at least one processor to perform operations to generate the target editable object from the data set.

15. The at least one non-transitory machine-readable medium of claim 13 , wherein the target editable object is an electronic chart that comprises a visualization of the data set.

16. The at lea

Assignees

• Microsoft Technology Licensing Llc

Inventors

• Wang Jinpeng
• Lin Chin-Yew

Classifications

• G06T11/26

  Drawing of charts or graphs · CPC title

• G06V30/40

  Document-oriented image-based pattern recognition · CPC title

• G06F40/177

  of tables; using ruled lines · CPC title

• G06F40/103 Primary

  Formatting, i.e. changing of presentation of documents (automatic justification G06F40/189; automatic line break hyphenation G06F40/191) · CPC title

• G06N3/08

  Learning methods · CPC title