18055752. EXTRACTING DOCUMENT HIERARCHY USING A MULTIMODAL, LAYER-WISE LINK PREDICTION NEURAL NETWORK simplified abstract (ADOBE INC.)

EXTRACTING DOCUMENT HIERARCHY USING A MULTIMODAL, LAYER-WISE LINK PREDICTION NEURAL NETWORK

Organization Name

ADOBE INC.

Inventor(s)

Vlad Morariu of Potomac MD (US)

Puneet Mathur of College Park MD (US)

Rajiv Jain of Vienna VA (US)

Ashutosh Mehra of Noida (IN)

Jiuxiang Gu of College Park MD (US)

Franck Dernoncourt of Sunnyvale CA (US)

Anandhavelu N of Kangayam (IN)

Quan Tran of San Jose CA (US)

Verena Kaynig-fittkau of Cambridge MA (US)

Nedim Lipka of Santa Clara CA (US)

Ani Nenkova of Philadelphia PA (US)

EXTRACTING DOCUMENT HIERARCHY USING A MULTIMODAL, LAYER-WISE LINK PREDICTION NEURAL NETWORK - A simplified explanation of the abstract

This abstract first appeared for US patent application 18055752 titled 'EXTRACTING DOCUMENT HIERARCHY USING A MULTIMODAL, LAYER-WISE LINK PREDICTION NEURAL NETWORK

Simplified Explanation

The present disclosure describes a system that generates a digital document hierarchy based on parent-child element relationships from visual elements. The system uses neural networks to classify elements and determine link probabilities, ultimately creating an interactive digital document.

• The system determines parent-child relationships from visual elements
• Utilizes neural networks to classify elements and determine link probabilities
• Generates a digital document hierarchy for interactive digital documents

Potential Applications

This technology could be applied in various fields such as web design, graphic design, digital publishing, and document management systems.

Problems Solved

This technology streamlines the process of creating digital documents by automatically organizing visual elements into a hierarchical structure, saving time and effort for designers and content creators.

Benefits

The benefits of this technology include improved efficiency in document creation, enhanced organization of visual elements, and the ability to create interactive digital documents with ease.

Potential Commercial Applications

Potential commercial applications of this technology include software tools for web designers, graphic designers, digital publishers, and document management systems providers.

Possible Prior Art

One possible prior art for this technology could be existing document management systems that offer some level of hierarchy for organizing digital documents. However, the use of neural networks for element classification and link probability determination may be a novel aspect of this innovation.

How does this technology handle complex visual elements with multiple parent-child relationships?

The system's ability to handle complex visual elements with multiple parent-child relationships is not explicitly addressed in the abstract. Further details on this aspect would provide a clearer understanding of the system's capabilities.

Can this technology be integrated with existing document creation software?

The abstract does not mention the compatibility of this technology with existing document creation software. Understanding how easily this system can be integrated with popular software tools would be beneficial for potential users looking to adopt this technology.

Original Abstract Submitted

The present disclosure relates to systems, methods, and non-transitory computer-readable media that generate a digital document hierarchy comprising layers of parent-child element relationships from the visual elements. For example, for a layer of the layers, the disclosed systems determine, from the visual elements, candidate parent visual elements and child visual elements. In addition, for the layer of the layers, the disclosed systems generate, from the feature embeddings utilizing a neural network, element classifications for the candidate parent visual elements and parent-child element link probabilities for the candidate parent visual elements and the child visual elements. Moreover, for the layer, the disclosed systems select parent visual elements from the candidate parent visual elements based on the parent-child element link probabilities. Further, the disclosed systems utilize the digital document hierarchy to generate an interactive digital document from the digital document image.