Adobe, Inc.

Adone Inc.

Duygu Ceylan Aksit

Jimei Yang

Yangtuanfeng Wang

Jingwan Lu

Zhili Chen

Krishna Kumar Singh

Giorgio Gori

Radomir Mech

Vladimir Kim

Mehmet Yumer

Niloy Jyoti Mitra

Jun Saito

Jose Ignacio Echevarria Vallespi

Xin Sun

Yijun Li

Tamy Boubekeur

Yi Zhou

Ruben Eduardo Villegas

Tobias Hinz

Jae Shin Yoon

Zhe Lin

Jianming Zhang

Nathan Aaron Carr

Kalyan Krishna Sunkavalli

Yang Zhou

David P Simons

Gavin S P Miller

Emiliano Gambaretto

Kevin Wampler

Daichi Ito

Jonathan Eisenmann

Noam Aigerman

Qing Liu

Siddhartha Chaudhuri

Matheus Gadelha

Tenell Rhodes

Eunbyung Park

Jeong Joon Park

Elya Shechtman

Hailin Jin

Baldo Antonio Faieta

Byungmoon Kim

Shabnam Ghadar

Jovan Popović

Geoffrey Oxholm

Srinivasa Madhava Phaneendra Angara

Kyle Olszewski

Alex Charles Filipkowski

Kuldeep Kulkarni

Gahye Park

Minhyuk Sung

Jinrong Xie

Radomir Meeh

Mianlun Zheng

Weiyue Wang

Rebecca Ilene Milman

Rinat Abdrashitov

Guilin Liu

Jingwei Huang

Pranjali Kokare

Elnaz Morad

Eric-Tuan Le

Yunseok Jang

Elya Shectman

Hugo Bertiche Argila

James Zachary

Chengan He

Chun Hao Huang

A system that utilizes neural networks to generate looping animations from still images. The system fits a 3D model to a pose of a person in a digital image. The system receives a 3D animation sequence that transitions between a starting pose and an ending pose. The system generates, utilizing an animation transition neural network, first and second 3D animation transition sequences that respectively transition between the pose of the person and the starting pose and between the ending pose and the pose of the person. The system modifies each of the 3D animation sequence, the first 3D animation transition sequence, and the second 3D animation transition sequence by applying a texture map. The system generates a looping 3D animation by combining the modified 3D animation sequence, the modified first 3D animation transition sequence, and the modified second 3D animation transition sequence.

Generating three-dimensional looping animations from still images

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.

Utilizing a warped digital image with a reposing model to synthesize a modified digital image

The present disclosure relates to systems, methods, and non-transitory computer-readable media that modify two-dimensional images via scene-based editing using three-dimensional representations of the two-dimensional images. For instance, in one or more embodiments, the disclosed systems utilize three-dimensional representations of two-dimensional images to generate and modify shadows in the two-dimensional images according to various shadow maps. Additionally, the disclosed systems utilize three-dimensional representations of two-dimensional images to modify humans in the two-dimensional images. The disclosed systems also utilize three-dimensional representations of two-dimensional images to provide scene scale estimation via scale fields of the two-dimensional images. In some embodiments, the disclosed systems utilizes three-dimensional representations of two-dimensional images to generate and visualize 3D planar surfaces for modifying objects in two-dimensional images. The disclosed systems further use three-dimensional representations of two-dimensional images to customize focal points for the two-dimensional images.

Generating three-dimensional human models representing two-dimensional humans in two-dimensional images

The present disclosure relates to systems, methods, and non-transitory computer readable media that utilizes neural networks to generate cinemagraphs from single RGB images. For example, the cyclic animation system includes a cyclic animation neural network trained with synthetic data, wherein different wind effects can be replicated using physically based simulations to create cyclic videos more efficiently. More specifically, the cyclic animation system generalizes a solution by operating in the gradient domain and using surface normal maps. Because normal maps are invariant to appearance (color, texture, illumination, etc.), the gap between synthetic and real data distribution in the normal map space is smaller than in the RGB space. The cyclic animation system performs a reshading approach that synthesizes RGB pixels from the original image and the animated normal maps to create plausible changes to the real image to create the cinemagraph.

Creating cinemagraphs from a single image

Modifying poses of two-dimensional humans in two-dimensional images by reposing three-dimensional human models representing the two-dimensional humans

Human inpainting utilizing a segmentation branch for generating an infill segmentation map

Generating a modified digital image utilizing a human inpainting model

The present disclosure relates to systems, non-transitory computer-readable media, and method that utilize a character animation neural network informed by motion and pose signatures to generate a digital video through person-specific appearance modeling and motion retargeting. In particular embodiments, the disclosed systems implement a character animation neural network that includes a pose embedding model to encode a pose signature into spatial pose features. The character animation neural network further includes a motion embedding model to encode a motion signature into motion features. In some embodiments, the disclosed systems utilize the motion features to refine per-frame pose features and improve temporal coherency. In certain implementations, the disclosed systems also utilize the motion features to demodulate neural network weights used to generate an image frame of a character in motion based on the refined pose features.

Generating animated digital videos utilizing a character animation neural network informed by pose and motion embeddings

Various disclosed embodiments are directed to estimating that a first vertex of a patch will change from a first position to a second position (the second position being at least partially indicative of secondary motion) based at least in part on one or more features of: primary motion data, one or more material properties, and constraint data associated with the particular patch. Such estimation can be made for some or all of the patches of an entire volumetric mesh in order to accurately predict the overall secondary motion of an object. This, among other functionality described herein resolves the inaccuracies, computer resource consumption, and the user experience of existing technologies.

Predicting secondary motion of multidimentional objects based on local patch features

Generative shape creation and editing is leveraged in a digital medium environment. An object editor system represents a set of training shapes as sets of visual elements known as 'handles,' and converts sets of handles into signed distance field (SDF) representations. A handle processor model is then trained using the SDF representations to enable the handle processor model to generate new shapes that reflect salient visual features of the training shapes. The trained handle processor model, for instance, generates new sets of handles based on salient visual features learned from the training handle set. Thus, utilizing the described techniques, accurate characterizations of a set of shapes can be learned and used to generate new shapes. Further, generated shapes can be edited and transformed in different ways.

Generative shape creation and editing

This disclosure describes methods, non-transitory computer readable storage media, and systems that generate realistic shading for three-dimensional objects inserted into digital images. The disclosed system utilizes a light encoder neural network to generate a representation embedding of lighting in a digital image. Additionally, the disclosed system determines points of the three-dimensional object visible within a camera view. The disclosed system generates a self-occlusion map for the digital three-dimensional object by determining whether fixed sets of rays uniformly sampled from the points intersects with the digital three-dimensional object. The disclosed system utilizes a generator neural network to determine a shading map for the digital three-dimensional object based on the representation embedding of lighting in the digital image and the self-occlusion map. Additionally, the disclosed system generates a modified digital image with the three-dimensional object inserted into the digital image with consistent lighting of the three-dimensional object and the digital image.

Inserting three-dimensional objects into digital images with consistent lighting via global and local lighting information

Embodiments of the present invention are directed towards intuitive editing of three-dimensional models. In embodiments, salient geometric features associated with a three-dimensional model defining an object are identified. Thereafter, feature attributes associated with the salient geometric features are identified. A feature set including a plurality of salient geometric features related to one another is generated based on the determined feature attributes (e.g., properties, relationships, distances). An editing handle can then be generated and displayed for the feature set enabling each of the salient geometric features within the feature set to be edited in accordance with a manipulation of the editing handle. The editing handle can be displayed in association with one of the salient geometric features of the feature set.

Intuitive editing of three-dimensional models

Embodiments provide systems, methods, and computer storage media for fitting 3D primitives to a 3D point cloud. In an example embodiment, 3D primitives are fit to a 3D point cloud using a global primitive fitting network that evaluates the entire 3D point cloud and a local primitive fitting network that evaluates local patches of the 3D point cloud. The global primitive fitting network regresses a representation of larger (global) primitives that fit the global structure. To identify smaller 3D primitives for regions with fine detail, local patches are constructed by sampling from a pool of points likely to contain fine detail, and the local primitive fitting network regresses a representation of smaller (local) primitives that fit the local structure of each of the local patches. The global and local primitives are merged into a combined, multi-scale set of fitted primitives, and representative primitive parameters are computed for each fitted primitive.

Fitting 3D primitives to a high-resolution point cloud

Motion retargeting with kinematic constraints is implemented in a digital medium environment. Generally, the described techniques provide for retargeting motion data from a source motion sequence to a target visual object. Accordingly, the described techniques position a target visual object in a defined visual environment to identify kinematic constraints of the target object relative to the visual environment. Further, the described techniques utilize an iterative optimization process that fine tunes the conformance of retargeted motion of a target object to the identified kinematic constraints.

Motion retargeting with kinematic constraints

The present disclosure relates to systems, methods, and non-transitory computer readable media for accurately, efficiently, and flexibly identifying and providing digital images of human figures in poses corresponding to a query pose. In particular, the disclosed systems can provide multiple approaches to searching for and providing pose images, including identifying a digital image depicting a human figure in a particular pose based on a query digital image that depicts the pose or identifying a digital image depicting a human figure in a particular pose based on a virtual mannequin. Indeed, the disclosed systems can provide a manipulable virtual mannequin that defines a query pose for searching a repository of digital images. Additionally, the disclosed systems can generate and provide digital pose image groups by clustering digital images together according to poses of human figures within a pose feature space.

Identifying and providing digital images depicting human poses utilizing visual interactive content search and virtual mannequins

Growing community of inventors

San Jose, CA, United States of America

Duygu Ceylan Aksit