US 10,762,135 B2Granted

Recommending software actions to create an image and recommending images to demonstrate the effects of software actions

AssigneeAdobe Inc.

InventorsMatthew Douglas Hoffman, Longqi Yang, Hailin Jin, Chen Fang

Life of this patent

Estimating the term…

Loading patent…

Abstract

A digital medium environment includes an asset processing application that performs editing of assets. A projection function is trained using pairs of actions pertaining to software edits, and assets resulting from the actions to learn a joint embedding between the actions and the assets. The projection function is used in the asset processing application to recommend software actions to create an asset, and also to recommend assets to demonstrate the effects of software actions. Recommendations are based on ranking distance measures that measure distances between actions representations and asset representations in a vector space.

Claims20

Dependent claims

In a digital medium environment including an asset processing application, a method implemented by at least one computing device via the asset processing application, the method comprising:

determining whether an input representation includes an actions vector or an asset vector, the asset vector representing content of an asset, the actions vector having elements that indicate a number of occurrences of software edits in a list of the software edits usable to edit the asset, the actions vector belonging to a vector space having dimensions corresponding to the software edits;
if the input representation is determined to include the actions vector:
ranking a plurality of asset representations according to a distance measure between the actions vector and the asset representations in the vector space;
and
recommending at least one of the plurality of asset representations based on the ranking of the plurality of asset representations;
and
if the input representation is determined to include the asset vector:
processing the asset vector with a projection function that projects the asset vector to the vector space to form a processed asset representation;
ranking a plurality of actions representations according to the distance measure between the processed asset representation and the plurality of actions representations in the vector space;
and
recommending a set of actions based on the ranking of the plurality of actions representations.

8 more elements

A system implemented by at least one computing device in a digital medium environment, the system comprising:

a processing system;
and
at least one computer readable medium storing instructions executable via the processing system to implement an asset processing application configured to:
obtain an asset representation including an asset vector derived from content of an asset;
access a plurality of actions representations each including an actions vector derived from a list of software edits usable to edit the asset, the actions vector having elements that indicate a number of occurrences of the software edits in the list of the software edits, the actions vector belonging to a vector space having dimensions corresponding to the software edits;
process the asset representation with a projection function by setting a length of the asset vector equal to a length of the actions vectors to form a processed asset representation;
rank the actions representations according to a distance measure between the processed asset representation and the actions vectors in the vector space;
and
recommend a set of actions based on the rank of the actions representations.

6 more elements

In a digital medium environment including an asset processing application, a method implemented by at least one computing device via the asset processing application, the method comprising:

a step for receiving a list of actions including software edits for editing assets;
a step for generating an actions representation by forming a vector based on the list of actions, the vector having elements that indicate a number of occurrences of the software edits in the list of actions, the vector belonging to a vector space having dimensions corresponding to the software edits;
a step for ranking a plurality of asset representations according to a distance measure between the actions representation and the plurality of asset representations, each asset representation of the plurality of asset representations having a dimensionality equal to a dimensionality of the actions representation determined by a length of the vector;
and
a step for recommending at least one of the asset representations based on the ranking of the plurality of asset representations.

2 more elements

Dependencies are inferred from the claim text.