Abstract: The present disclosure describes a model for large scale color prediction of objects identified in images. Embodiments of the present disclosure include an object detection network, an attention network, and a color classification network. The object detection network generates object features for an object in an image and may include a convolutional neural network (CNN), region proposal network, or a ResNet. The attention network generates an attention vector for the object based on the object features, wherein the attention network takes a query vector based on the object features, and a plurality of key vector and a plurality of value vectors corresponding to a plurality of colors as input. The color classification network generates a color attribute vector based on the attention vector, wherein the color attribute vector indicates a probability of the object including each of the plurality of colors.

Abstract: A curve sampling technique for generating transformed digital visual content is leveraged in a digital medium environment. Initially, a curve sampling system obtains digital visual content, e.g., images and videos. The curve sampling system generates transformed digital visual content by transforming one or more pixels of the digital visual content using a lookup table that is derived from samples of a curve taken at evenly spaced intervals along a y-axis of a graph of the curve. Broadly speaking, the curve defines how to transform a visual characteristic of the pixels in order to achieve a desired digital visual content transformation. Additionally, the curve sampling may correspond to one step in a series of steps for transforming colors of digital visual content. Indeed, such transformations may involve multiple curve sampling steps.

Abstract: The present disclosure describes multi-stage image editing techniques to improve detail and accuracy in edited images. An input image including a target region to be edited and an edit parameter specifying a modification to the target region are received. A parsing map of the input image is generated. A latent representation of the parsing map is generated. An edit is applied to the latent representation of the parsing map based on the edit parameter. The edited latent representation is input to a neural network to generate a modified parsing map including the target region with a shape change according to the edit parameter. Based on the input image and the modified parsing map, a masked image corresponding to the shape change is generated. Based on the masked image, a neural network is used to generate an edited image with the modification to the target region.

Abstract: Techniques and systems are provided for configuring neural networks to perform certain image manipulation operations. For instance, in response to obtaining an image for manipulation, an image manipulation system determines the fitness scores for a set of neural networks resulting from the processing of a noise map. Based on these fitness scores, the image manipulation system selects a subset of the set of neural networks for cross-breeding into a new generation of neural networks. The image manipulation system evaluates the performance of this new generation of neural networks and continues cross-breeding this neural networks until a fitness threshold is satisfied. From the final generation of neural networks, the image manipulation system selects a neural network that provides a desired output and uses the neural network to generate the manipulated image.

Abstract: In implementations of systems for generating templates using structure-based matching, a computing device implements a template system to receive input data describing a set of digital design elements. The template system represents the input data as a sentence in a design structure language that describes structural relationships between design elements included in the set of digital design elements. An input template embedding is generated based on the sentence in the design structure language. The template system generates a digital template that includes the set of digital design elements for display in a user interface based on the input template embedding.

Abstract: Offset object alignment operations are described that support an ability to control alignment operations to aid positioning of an object in relation to at least one other object in a user interface based an offset value. This is performable through identification of objects that overlap along an axis in a user interface and calculation of offset values using these object pairs. Filtering and priority based techniques are also usable as part of calculated an offset value to be used as part of an alignment operation.

Abstract: Constrained stroke editing techniques for digital content are described. In these examples, a stroke constraint system is employed as part of a digital content creation system to manage input, editing, and erasure (i.e., removal) of strokes via a user interface as part of editing digital content. To do so, locations and attributes of a displayed stroke are used to constrain location and/or attributes of an input stroke.

Abstract: A missing glyph replacement system is described. In an example, a Unicode identifier of a missing glyph is obtained and glyph metadata describing a glyph cluster that includes the Unicode identifier is obtained from a cache maintained in the storage device, e.g., as part of preprocessing. From this, the system obtains glyphs from the font using Unicode identifiers included in the glyph cluster. The system uses a representative glyph from these glyphs to verify the glyph cluster, and if verified obtains glyphs based on the cluster. For these obtained glyphs, an amount of similarity is determined for the missing glyph with respect to the plurality of obtained glyphs, e.g., to control output of representations of the obtained glyphs in the user interface. The representations are user selectable via the user interface to replace the missing glyph.

Abstract: An appearance-responsive material map generation system generates a set of material maps based on the appearance of a material depicted in the source material data. A neural network included in the appearance-responsive material map generation system is trained to identify features of particular source material data, such as features that contribute to a highly realistic appearance of a graphical object rendered with the material depicted in the source material data. In some cases, the trained neural network receives source material data that includes at least one source material map. Based on the features that are identified for the particular source material data, the appearance-responsive material map generation system creates a respective set of appearance-responsive material maps for the particular source material data. In some cases, the appearance-responsive rendering map set is arranged as an inconsistent pyramid of material maps.

Abstract: Systems and methods for customer journey optimization in email marketing are described. The systems and methods may identify a plurality of messages for a first time period, wherein the plurality of messages are categorized according to a plurality of messages types, identify user information for a customer, wherein the user information includes user interaction data, determine a message type from the plurality of message types for the first time period based on the user information, wherein the message type is determined using a decision making model comprising a deep Q-learning neural network, select a message from the plurality of messages based on the determined message type, and transmit the message to the customer during the first time period based on the selection.

Abstract: Techniques and systems are provided for training a machine learning model using different datasets to perform one or more tasks. The machine learning model can include a first sub-module configured to perform a first task and a second sub-module configured to perform a second task. The first sub-module can be selected for training using a first training dataset based on a format of the first training dataset. The first sub-module can then be trained using the first training dataset to perform the first task. The second sub-module can be selected for training using a second training dataset based on a format of the second training dataset. The second sub-module can then be trained using the second training dataset to perform the second task.

Abstract: The present disclosure relates to systems, non-transitory computer-readable media, and methods that generate a dynamic cross-platform ask interface and utilize a cross-platform language processing model to provide platform-specific, contextually based responses to natural language digital text queries. In particular, in one or more embodiments, the disclosed systems utilize machine learning models to extract registered intents from digital text queries to identify platform-specific configurations associated with the registered intents. Utilizing the platform-specific configurations, the disclosed systems can generate tailored platform-specific requests for information, as well as customized end-user search results that cause client devices to efficiently, accurately, and flexibly render platform-specific search results.

Abstract: Some techniques described herein relate to generating a hierarchical topic model (HTM), which can be used to generate custom content. In one example, a method includes determining first-level topics in a topic hierarchy related to a corpus of documents. A first-level topic of the first-level topics includes multiple words. The multiple words are grouped into clusters based on word embeddings of the multiple words. The multiple words are then subdivided into second-level topics as subtopics of the first-level topic, such that the number of second-level topics equals the number of clusters. A document of the corpus of documents is assigned to the first-level topic and to a second-level topic of the second-level topics, and an indication is received of access by a user to the document. Custom content is generated for the user based on one or more other documents assigned to the first-level topic and the second-level topic.

Abstract: Embodiments are disclosed for removing typographic rivers from electronic documents. The method may include receiving an electronic document including a plurality of words for automatic typographic correction. A typographic river is identified in the electronic document, the typographic river including a plurality of nodes, each node including an empty glyph. A candidate adjustment that removes the first node of the plurality of nodes is identified and the candidate adjustment is applied to the electronic document.

Abstract: Digital image text editing techniques as implemented by an image processing system are described that support increased user interaction in the creation and editing of digital images through understanding a content creator's intent as expressed using text. In one example, a text user input is received by a text input module. The text user input describes a visual object and a visual attribute, in which the visual object specifies a visual context of the visual attribute. A feature representation generated by a text-to-feature system using a machine-learning module based on the text user input. The feature representation is passed to an image editing system to edit a digital object in a digital image, e.g., by applying a texture to an outline of the digital object within the digital image.

Abstract: Self-consumable portions generation techniques from a digital document are described. The self-consumable portions are generated based on a determination of an amount of resources available at a receiver device that is to receive the digital document. Examples of the resources include an amount of memory resources, processing resources, and/or network resources associated with the receiver device. The self-consumable portions, once generated, are separately renderable at the receiver device.

Abstract: In implementations of systems for predicting a terminal event, a computing device implements a termination system to receive input data defining a period of time and a maximum event threshold. This system uses a classification model to generate event scores for a plurality of entity devices. Each of the event scores indicates a probability of an event occurrence for a corresponding entity device within a period of time. The plurality of entity devices are segmented into a first segment and a second segment based on an event score threshold. Entity devices included in the first segment have event scores greater than the event score threshold and entity devices included in the second segment have event scores below the event score threshold. The termination system generates an indication of a probability that a number of event occurrences for the entity devices included in the second segment exceeds the maximum even threshold within the period of time.

Abstract: Embodiments are disclosed for generating an intelligent change summary are described. In some embodiments, a method of generating an intelligent change summary includes obtaining a representation of a plurality of versions of a document, determining a distance score based on a comparison of a first of version of the document and a second version of the document, the distance score representing a magnitude of changes made from the first version of the document to the second version of the document, and generating a change summary of the document based on the distance score.

Abstract: The present disclosure includes methods and systems for generating digital predictive models by progressively sampling a repository of data samples. In particular, one or more embodiments of the disclosed systems and methods identify initial attributes for predicting a target attribute and utilize the initial attributes to identify a coarse sample set. Moreover, the disclosed systems and methods can utilize the coarse sample set to identify focused attributes pertinent to predicting the target attribute. Utilizing the focused attributes, the disclosed systems and methods can identify refined data samples and utilize the refined data samples to identify final attributes and generate a digital predictive model.

Abstract: A method for recommending digital content includes: determining user preferences and a time horizon of a given user; determining a group for the given user based on the determined user preferences; determining a number of users of the determined group and a similarity of the users; applying information including the number of users, the similarity, and the time horizon to a model selection classifier to select one of a personalized model of the user and a group model of the determined group; and running the selected model to determine digital content to recommend.