Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training a neural radiance field (NeRF) model on unposed images. In particular, the training incorporates a geometric consistency loss to train the encoder neural network that predicts the poses of the unposed images.

Abstract: A method comprising: receiving a reference facial image of a first subject, wherein the reference image represents a specified makeup style applied to a face of the first subject; receiving a target facial image of a target subject without makeup; performing pixel-wise alignment of the reference image to the target image; generating a translation of the reference image to obtain a de-makeup version of the reference image representing the face of the first subject without the specified makeup style; calculating an appearance modification contribution representing a difference between the reference image and the de-makeup version; and adding the calculated appearance modification contribution to the target image, to construct a modified the target image which represents the specified makeup style applied to a face of the target subject.

Abstract: A process and related system and computer program product for constructing a personalized contacting interface for a facial mask, comprising the steps of: providing a three-dimensional (3D) reference model representative of a human face; identifying in the reference model a desired contact area circumscribing one or more facial regions; generating a digital design model of a contacting interface, the digital design model having a perimeter configured to provide a continuous air seal along the desired contact area of the reference model; receiving a 3D facial target model corresponding to the face of a subject; performing an elastic transformation of the reference model to conform the reference model to the target model; modifying said perimeter of said digital design model based on the deformed reference model; and using said modified digital design model to generate a set of manufacturing instructions for a contacting interface personalized for said subject.

Abstract: A method that uses at least one hardware processor for receiving a three-dimensional model of an object, receiving an audio sequence embodied as a digital file that comprises a musical composition, generating a video frame sequence, and synthesizing the audio sequence and the video frame sequence into an audiovisual clip. The three-dimensional model is embodied as a digital file that comprises a representation of the object. The generating step comprises computing a caricature of the object by applying a computerized caricaturization algorithm to the three-dimensional model. The computing has scaling gradient fields of surface coordinates of the three-dimensional model by a function of a Gaussian curvature of the surface, and finding a regular surface whose gradient fields fit the scaled gradient fields.

Abstract: A method that uses at least one hardware processor for receiving a three-dimensional model of an object, receiving an audio sequence embodied as a digital file that comprises a musical composition, generating a video frame sequence, and synthesizing the audio sequence and the video frame sequence into an audiovisual clip. The three-dimensional model is embodied as a digital file that comprises a representation of the object. The generating step comprises computing a caricature of the object by applying a computerized caricaturization algorithm to the three-dimensional model. The computing has scaling gradient fields of surface coordinates of the three-dimensional model by a function of a Gaussian curvature of the surface, and finding a regular surface whose gradient fields fit the scaled gradient fields.

Abstract: Computerized, audio-based caricature exaggeration which includes: receiving a three-dimensional model of an object; receiving an audio sequence; generating a video frame sequence, said generating comprising computing a caricature of the object, wherein (a) the computing is with a different exaggeration factor for each of multiple ones of the video frames, and (b) the different exaggeration factor is based on one or more parameters of the audio sequence; and synthesizing the audio sequence and the video frame sequence into an audiovisual clip.

Abstract: Computerized, audio-based caricature exaggeration which includes: receiving a three-dimensional model of an object; receiving an audio sequence; generating a video frame sequence, said generating comprising computing a caricature of the object, wherein (a) the computing is with a different exaggeration factor for each of multiple ones of the video frames, and (b) the different exaggeration factor is based on one or more parameters of the audio sequence; and synthesizing the audio sequence and the video frame sequence into an audiovisual clip.