18190556. HUMAN INPAINTING UTILIZING A SEGMENTATION BRANCH FOR GENERATING AN INFILL SEGMENTATION MAP simplified abstract (Adobe Inc.)

HUMAN INPAINTING UTILIZING A SEGMENTATION BRANCH FOR GENERATING AN INFILL SEGMENTATION MAP

Organization Name

Adobe Inc.

Inventor(s)

Krishna Kumar Singh of San Jose CA (US)

Yijun Li of Seattle WA (US)

Jingwan Lu of Santa Clara CA (US)

Duygu Ceylan Aksit of Mountain View CA (US)

Yangtuanfeng Wang of London (GB)

Jimei Yang of Merced CA (US)

Tobias Hinz of Ulm (DE)

Qing Liu of Santa Clara CA (US)

Jianming Zhang of Campbell CA (US)

Zhe Lin of Fremont CA (US)

HUMAN INPAINTING UTILIZING A SEGMENTATION BRANCH FOR GENERATING AN INFILL SEGMENTATION MAP - A simplified explanation of the abstract

This abstract first appeared for US patent application 18190556 titled 'HUMAN INPAINTING UTILIZING A SEGMENTATION BRANCH FOR GENERATING AN INFILL SEGMENTATION MAP

The present disclosure involves systems, methods, and computer-readable media that use artificial intelligence to edit digital images based on scene understanding.

• Utilizes generative machine learning models to modify digital images depicting human subjects.
• Performs infill modifications or human inpainting to complete or enhance digital images.
• Reposes subjects within digital images to create modified versions.
• Conducts facial expression transfer and animations to alter digital images or create animations.

Potential Applications: This technology can be used in the fields of photography, graphic design, and entertainment industries for enhancing and manipulating digital images.

Problems Solved: This technology addresses the need for efficient and effective editing of digital images, especially those portraying human subjects, by utilizing AI-based image understanding.

Benefits: The benefits of this technology include improved image editing capabilities, enhanced creativity in digital image manipulation, and the ability to generate realistic modifications.

Commercial Applications: "AI-Powered Digital Image Editing Technology for Enhanced Visual Content Creation"

This technology can be utilized by graphic design companies, photography studios, and entertainment production houses to streamline the image editing process and create visually appealing content.

Prior Art: There are existing technologies for digital image editing, but the use of AI for scene-based editing and human subject manipulation is a novel approach.

Frequently Updated Research: Research on the advancements in generative machine learning models for image editing and AI-based scene understanding is ongoing in the field of computer vision.

Questions about AI-Powered Digital Image Editing Technology:

Question 1: How does this technology differ from traditional methods of digital image editing? Answer: This technology leverages artificial intelligence and machine learning to understand and manipulate digital images, providing more advanced and efficient editing capabilities compared to traditional methods.

Question 2: What are the potential ethical considerations associated with using AI for editing human subjects in digital images? Answer: Ethical considerations may arise regarding consent, privacy, and the potential misuse of AI-generated images, highlighting the importance of responsible use and regulation in this field.

Original Abstract Submitted

The present disclosure relates to systems, methods, and non-transitory computer-readable media that modify digital images via scene-based editing using image understanding facilitated by artificial intelligence. For example, in one or more embodiments the disclosed systems utilize generative machine learning models to create modified digital images portraying human subjects. In particular, the disclosed systems generate modified digital images by performing infill modifications to complete a digital image or human inpainting for portions of a digital image that portrays a human. Moreover, in some embodiments, the disclosed systems perform reposing of subjects portrayed within a digital image to generate modified digital images. In addition, the disclosed systems in some embodiments perform facial expression transfer and facial expression animations to generate modified digital images or animations.