Provided are a face image processing method and apparatus, an image device, and a storage medium. The face image processing method includes: acquiring first-key-point information of a first face image; performing position transformation on the first-key-point information to obtain second-key-point information conforming to a second facial geometric attribute, the second facial geometric attribute being different from a first facial geometric attribute corresponding to the first-key-point information; and performing facial texture coding processing by utilizing a neural network and the second-key-point information to obtain a second face image.

Illustrative systems and methods for determining and using three-dimensional (3D) meshes of a subject are described herein. An illustrative system obtains 3D meshes of a subject at a frame rate, the 3D meshes including a current-frame 3D mesh corresponding to a current frame and a previous-frame 3D mesh corresponding to a previous frame. The system deforms the previous-frame 3D mesh based on the current-frame 3D mesh and determines, based on the deformed previous-frame 3D mesh, a 3D mesh of the subject for use with the current frame. In certain examples, the system uses the 3D mesh of the subject to facilitate rendering of a viewpoint-adaptive 3D representation of the subject.

Systems and methods relate receiving video streams captured of a subject by video cameras, each video stream including video frames that are time-synchronized with the video, each video camera having a known vantage point in a predetermined coordinate system; obtaining at least one three-dimensional (3D) mesh of the subject, the mesh being time-synchronized and including a plurality of mesh vertices with known locations; identifying a user-selected viewpoint, and identifying a viewpoint-specific subset of the mesh vertices visible; generating 3D submeshes of the subject by calculating visible-vertices lists from the vantage point of each video camera from which the viewpoint-specific subset of mesh vertices is visible; projecting mesh vertices from the calculated visible-vertices lists on to video pixels; and rendering viewpoint-adaptive 3D personas of the subject by weighting video pixel colors from different video-camera vantage points according to the geometric relationship of each video-camera vantage point to the user-selected viewpoint.

Systems and methods relate to encoded video streams including geometric-data streams transmitted to a receiver for rendering of a viewpoint-adaptive 3D persona. A method includes obtaining a three-dimensional (3D) mesh of a subject generated from depth-camera-captured information about the subject, obtaining a facial-mesh model, locating a facial portion of the obtained 3D mesh of the subject, computing a geometric transform based on the facial portion and the facial-mesh model, the geometric transform determined in response to one or more aggregated error differences between a plurality of feature points on the facial-mesh model and a plurality of corresponding feature points on the facial portion of the obtained 3D mesh, generating a transformed facial-mesh model using the geometric transform and generating a hybrid mesh of the subject at least in part by combining the transformed facial-mesh model and at least a portion of the obtained 3D mesh.

Systems and methods relate to receiving a plurality of video streams captured of a subject by a plurality of video cameras, each video stream including video frames time-synchronized according to a shared frame rate, each video camera having a known vantage point in a predetermined coordinate system; obtaining at least one three-dimensional (3D) mesh of the subject at the shared frame rate, the 3D mesh time-synchronized with the video frames of the video streams, the at least one mesh including a plurality of vertices with known locations in the predetermined coordinate system; calculating one or more lists of visible-vertices at the shared frame rate, each list including a subset of the plurality of vertices of the at least one 3D mesh of the subject, the subset being a function of the location of the known vantage point associated with at least one of the plurality of video cameras; generating one or more time-synchronized data streams at the shared frame rate, the one or more time-synchronized data streams including: one or more video streams encoding at least one of the plurality of video streams; and one or more geometric-data streams including the calculated one or more visible-vertices lists; and transmitting the one or more time-synchronized data streams to a receiver for rendering of a viewpoint-adaptive 3D persona of the subject.

An exemplary method includes generating a 3D mesh of a subject based on frames of time-synchronized video streams of a subject, the frames associated with a first time and generating a transformed facial-mesh model based on a facial portion of the 3D mesh and a facial-mesh model. The method further includes generating a hybrid mesh by combining the transformed facial-mesh model and at least a portion of the 3D mesh. The method further includes generating a current 3D mesh based on frames of the time-synchronized video streams associated with a second time that temporally follows the first time. The method further includes generating a deformed historical 3D mesh by applying a non-rigid deformation process to the hybrid mesh based on the current 3D mesh. The method further includes compressing the deformed historical 3D mesh to form at least one triangle-based 3D submesh including a plurality of submesh triangles.

The disclosure relates to systems, methods and programs for geometrically constrained, unsupervised training of convolutional autoencoders on unlabeled images for extracting eye landmarks.

An illustrative method includes obtaining a three-dimensional (3D) mesh of a subject, obtaining a mesh model, and generating a hybrid mesh of the subject. The generating includes replacing a portion of the 3D mesh with the mesh model such that the hybrid mesh includes a non-replaced portion of the 3D mesh represented at a first resolution and the mesh model representing the replaced portion of the 3D mesh at a second resolution.

Disclosed is a computer-implemented method of determining one or more position and/or orientation parameters of an anatomical structure of a body portion. The anatomical structure has a longitudinal shape defining a longitudinal axis. The method includes generating and/or reading, by a data processing system, volumetric data of at least a portion of a subject. The method further includes generating and/or reading, by the data processing system, a deformable template which provides an estimate for a location of the longitudinal axis in the portion of the subject. The method further includes matching, by the data processing system, the deformable template to the volumetric data, thereby obtaining a matched template. The matching comprises using one or more internal energy functions and one or more external energy functions for optimizing an objective function. The method further includes determining, by the data processing system, the at least one position and/or orientation parameter based on the matched template.

Methods for determining metrology sites for products includes detecting corresponding objects in measurement data of one or more product samples, and aligning the detected objects are aligned. The methods also include analyzing the aligned objects, and determining metrology sites based on the analysis. Devices use such methods to determine metrology sites for products.