Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for iteratively decimating a three-dimensional mesh utilizing successive self-parameterization. For example, the disclosed system can self-parameterize local geometries of a three-dimensional mesh using surface mappings within a two-dimensional surface mapping space. The disclosed system can collapse edges in the three-dimensional mesh to create new vertices from the collapsed edges. The disclosed system can parameterize the collapsed edges based on the surface mappings to collapse corresponding edges within the surface mapping space. The disclosed system can thus generate a decimated three-dimensional mesh by collapsing edges in the three-dimensional mesh while providing a bijective map between points in the decimated three-dimensional mesh and corresponding points in the three-dimensional mesh.

Abstract: Systems and methods for performing image search are described. An image search method may include generating a feature vector for each of a plurality of stored images using a machine learning model trained using a rotation loss term, receiving a search query comprising a search image with object having an orientation, generating a query feature vector for the search image using the machine learning model, wherein the query feature vector is based at least in part on the orientation, comparing the query feature vector to the feature vector for each of the plurality of stored images, and selecting at least one stored image of the plurality of stored images based on the comparison, wherein the at least one stored image comprises a similar orientation to the orientation of the object in the search image.

Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for reconstructing three-dimensional meshes from two-dimensional images of objects with automatic coordinate system alignment. For example, the disclosed system can generate feature vectors for a plurality of images having different views of an object. The disclosed system can process the feature vectors to generate coordinate-aligned feature vectors aligned with a coordinate system associated with an image. The disclosed system can generate a combined feature vector from the feature vectors aligned to the coordinate system. Additionally, the disclosed system can then generate a three-dimensional mesh representing the object from the combined feature vector.

Abstract: Certain embodiments detect human-object interactions in image content. For example, human-object interaction metadata is applied to an input image, thereby identifying contact between a part of a depicted human and a part of a depicted object. Applying the human-object interaction metadata involves computing a joint-location heat map by applying a pose estimation subnet to the input image and a contact-point heat map by applying an object contact subnet to the to the input image. The human-object interaction metadata is generated by applying an interaction-detection subnet to the joint-location heat map and the contact-point heat map. The interaction-detection subnet is trained to identify an interaction based on joint-object contact pairs, where a joint-object contact pair includes a relationship between a human joint location and a contact point. An image search system or other computing system is provided with access to the input image having the human-object interaction metadata.

Abstract: Certain embodiments involve techniques for generating a 3D representation based on a provided 2D image of an object. An image generation system receives the 2D image representation and generates a multi-dimensional vector of the input that represents the image. The image generation system samples a set of points and provides the set of points and the multi-dimensional vector to a neural network that was trained to predict a 3D surface representing the image such that the 3D surface is consistent with a 3D surface of the object calculated using an implicit function for representing the image. The neural network predicts, based on the multi-dimensional vector and the set of points, the 3D surface representing the object.

Abstract: 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.

Abstract: In implementations of object animation using generative neural networks, one or more computing devices of a system implement an animation system for reproducing animation of an object in a digital video. A mesh of the object is obtained from a first frame of the digital video and a second frame of the digital video having the object is selected. Features of the object from the second frame are mapped to vertices of the mesh, and the mesh is warped based on the mapping. The warped mesh is rendered as an image by a neural renderer and compared to the object from the second frame to train a neural network. The rendered image is then refined by a generator of a generative adversarial network which includes a discriminator. The discriminator trains the generator to reproduce the object from the second frame as the refined image.

Abstract: 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.

Abstract: Matching an illumination of an embedded virtual object (VO) with current environment illumination conditions provides an enhanced immersive experience to a user. To match the VO and environment illuminations, illumination basis functions are determined based on preprocessing image data, captured as a first combination of intensities of direct illumination sources illuminates the environment. Each basis function corresponds to one of the direct illumination sources. During the capture of runtime image data, a second combination of intensities illuminates the environment. An illumination-weighting vector is determined based on the runtime image data. The determination of the weighting vector accounts for indirect illumination sources, such as surface reflections. The weighting vector encodes a superposition of the basis functions that corresponds to the second combination of intensities. The method illuminates the VO based on the weighting vector.

Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for reconstructing three-dimensional object meshes from two-dimensional images of objects using multi-view cycle projection. For example, the disclosed system can determine a multi-view cycle projection loss across a plurality of images of an object via an estimated three-dimensional object mesh of the object. For example, the disclosed system uses a pixel mapping neural network to project a sampled pixel location across a plurality of images of an object and via a three-dimensional mesh representing the object. The disclosed system determines a multi-view cycle consistency loss based on a difference between the sampled pixel location and a cycle projection of the sampled pixel location and uses the loss to update the pixel mapping neural network, a latent vector representing the object, or a shape generation neural network that uses the latent vector to generate the object mesh.

Abstract: Techniques are described herein for generating and using a unified shape representation that encompasses features of different types of shape representations. In some embodiments, the unified shape representation is a unicode comprising a vector of embeddings and values for the embeddings. The embedding values are inferred, using a neural network that has been trained on different types of shape representations, based on a first representation of a three-dimensional (3D) shape. The first representation is received as input to the trained neural network and corresponds to a first type of shape representation. At least one embedding has a value dependent on a feature provided by a second type of shape representation and not provided by the first type of shape representation. The value of the at least one embedding is inferred based upon the first representation and in the absence of the second type of shape representation for the 3D shape.

Abstract: Techniques are disclosed for 3D object reconstruction using photometric mesh representations. A decoder is pretrained to transform points sampled from 2D patches of representative objects into 3D polygonal meshes. An image frame of the object is fed into an encoder to get an initial latent code vector. For each frame and camera pair from the sequence, a polygonal mesh is rendered at the given viewpoints. The mesh is optimized by creating a virtual viewpoint, rasterized to obtain a depth map. The 3D mesh projections are aligned by projecting the coordinates corresponding to the polygonal face vertices of the rasterized mesh to both selected viewpoints. The photometric error is determined from RGB pixel intensities sampled from both frames. Gradients from the photometric error are backpropagated into the vertices of the assigned polygonal indices by relating the barycentric coordinates of each image to update the latent code vector.

Abstract: The present disclosure includes methods and systems for identifying and manipulating a segment of a three-dimensional digital model based on soft classification of the three-dimensional digital model. In particular, one or more embodiments of the disclosed systems and methods identify a soft classification of a digital model and utilize the soft classification to tune segmentation algorithms. For example, the disclosed systems and methods can utilize a soft classification to select a segmentation algorithm from a plurality of segmentation algorithms, to combine segmentation parameters from a plurality of segmentation algorithms, and/or to identify input parameters for a segmentation algorithm. The disclosed systems and methods can utilize the tuned segmentation algorithms to accurately and efficiently identify a segment of a three-dimensional digital model.

Abstract: This disclosure relates to methods, non-transitory computer readable media, and systems that use style-aware puppets patterned after a source-character-animation sequence to generate a target-character-animation sequence. In particular, the disclosed systems can generate style-aware puppets based on an animation character drawn or otherwise created (e.g., by an artist) for the source-character-animation sequence. The style-aware puppets can include, for instance, a character-deformational model, a skeletal-difference map, and a visual-texture representation of an animation character from a source-character-animation sequence. By using style-aware puppets, the disclosed systems can both preserve and transfer a detailed visual appearance and stylized motion of an animation character from a source-character-animation sequence to a target-character-animation sequence.

Abstract: Techniques are disclosed for 3D object reconstruction using photometric mesh representations. A decoder is pretrained to transform points sampled from 2D patches of representative objects into 3D polygonal meshes. An image frame of the object is fed into an encoder to get an initial latent code vector. For each frame and camera pair from the sequence, a polygonal mesh is rendered at the given viewpoints. The mesh is optimized by creating a virtual viewpoint, rasterized to obtain a depth map. The 3D mesh projections are aligned by projecting the coordinates corresponding to the polygonal face vertices of the rasterized mesh to both selected viewpoints. The photometric error is determined from RGB pixel intensities sampled from both frames. Gradients from the photometric error are backpropagated into the vertices of the assigned polygonal indices by relating the barycentric coordinates of each image to update the latent code vector.

Abstract: In implementations of object animation using generative neural networks, one or more computing devices of a system implement an animation system for reproducing animation of an object in a digital video. A mesh of the object is obtained from a first frame of the digital video and a second frame of the digital video having the object is selected. Features of the object from the second frame are mapped to vertices of the mesh, and the mesh is warped based on the mapping. The warped mesh is rendered as an image by a neural renderer and compared to the object from the second frame to train a neural network. The rendered image is then refined by a generator of a generative adversarial network which includes a discriminator. The discriminator trains the generator to reproduce the object from the second frame as the refined image.

Abstract: The present disclosure includes methods and systems for identifying and manipulating a segment of a three-dimensional digital model based on soft classification of the three-dimensional digital model. In particular, one or more embodiments of the disclosed systems and methods identify a soft classification of a digital model and utilize the soft classification to tune segmentation algorithms. For example, the disclosed systems and methods can utilize a soft classification to select a segmentation algorithm from a plurality of segmentation algorithms, to combine segmentation parameters from a plurality of segmentation algorithms, and/or to identify input parameters for a segmentation algorithm. The disclosed systems and methods can utilize the tuned segmentation algorithms to accurately and efficiently identify a segment of a three-dimensional digital model.

Abstract: Matching an illumination of an embedded virtual object (VO) with current environment illumination conditions provides an enhanced immersive experience to a user. To match the VO and environment illuminations, illumination basis functions are determined based on preprocessing image data, captured as a first combination of intensities of direct illumination sources illuminates the environment. Each basis function corresponds to one of the direct illumination sources. During the capture of runtime image data, a second combination of intensities illuminates the environment. An illumination-weighting vector is determined based on the runtime image data. The determination of the weighting vector accounts for indirect illumination sources, such as surface reflections. The weighting vector encodes a superposition of the basis functions that corresponds to the second combination of intensities. The method illuminates the VO based on the weighting vector.

Abstract: 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.

Abstract: Techniques are disclosed for the synthesis of a full set of slotted content, based upon only partial observations of the slotted content. With respect to a font, the slots may comprise particular letters or symbols or glyphs in an alphabet. Based upon partial observations of a subset of glyphs from a font, a full set of the glyphs corresponding to the font may be synthesized and may further be ornamented.