Abstract: Example embodiments relate to techniques for volumetric performance capture with neural rendering. A technique may involve initially obtaining images that depict a subject from multiple viewpoints and under various lighting conditions using a light stage and depth data corresponding to the subject using infrared cameras. A neural network may extract features of the subject from the images based on the depth data and map the features into a texture space (e.g., the UV texture space). A neural renderer can be used to generate an output image depicting the subject from a target view such that illumination of the subject in the output image aligns with the target view. The neural render may resample the features of the subject from the texture space to an image space to generate the output image.

Abstract: Apparatus and methods related to applying lighting models to images are provided. An example method includes receiving, via a computing device, an image comprising a subject. The method further includes relighting, via a neural network, a foreground of the image to maintain a consistent lighting of the foreground with a target illumination. The relighting is based on a per-pixel light representation indicative of a surface geometry of the foreground. The light representation includes a specular component, and a diffuse component, of surface reflection. The method additionally includes predicting, via the neural network, an output image comprising the subject in the relit foreground. One or more neural networks can be trained to perform one or more of the aforementioned aspects.

Abstract: Techniques of image synthesis using a neural radiance field (NeRF) includes generating a deformation model of movement experienced by a subject in a non-rigidly deforming scene. For example, when an image synthesis system uses NeRFs, the system takes as input multiple poses of subjects for training data. In contrast to conventional NeRFs, the technical solution first expresses the positions of the subjects from various perspectives in an observation frame. The technical solution then involves deriving a deformation model, i.e., a mapping between the observation frame and a canonical frame in which the subject's movements are taken into account. This mapping is accomplished using latent deformation codes for each pose that are determined using a multilayer perceptron (MLP). A NeRF is then derived from positions and casted ray directions in the canonical frame using another MLP. New poses for the subject may then be derived using the NeRF.

Abstract: A wearable computing device includes a frame, a camera mounted on the frame so as to capture images of an environment outside of the wearable computing device, a display device mounted on the frame so as to display the images captured by the camera, and at least one eye gaze tracking device mounted on the frame so as to track a gaze directed at the images displayed by the display device. In response to the detection of a fixation of the gaze on the display of images, the system may identify a pixel area corresponding to a fixation point of the fixation gaze on the display of images. The system may identify an object in the ambient environment corresponding to the identified pixel area, and set the identified object as a selected object for user interaction.

Abstract: Apparatus and methods related to applying lighting models to images of objects are provided. An example method includes applying a geometry model to an input image to determine a surface orientation map indicative of a distribution of lighting on an object based on a surface geometry. The method further includes applying an environmental light estimation model to the input image to determine a direction of synthetic lighting to be applied to the input image. The method also includes applying, based on the surface orientation map and the direction of synthetic lighting, a light energy model to determine a quotient image indicative of an amount of light energy to be applied to each pixel of the input image. The method additionally includes enhancing, based on the quotient image, a portion of the input image. One or more neural networks can be trained to perform one or more of the aforementioned aspects.

Abstract: Example embodiments relate to techniques for volumetric performance capture with neural rendering. A technique may involve initially obtaining images that depict a subject from multiple viewpoints and under various lighting conditions using a light stage and depth data corresponding to the subject using infrared cameras. A neural network may extract features of the subject from the images based on the depth data and map the features into a texture space (e.g., the UV texture space). A neural renderer can be used to generate an output image depicting the subject from a target view such that illumination of the subject in the output image aligns with the target view. The neural render may resample the features of the subject from the texture space to an image space to generate the output image.

Abstract: Systems and methods are directed to encoding and/or decoding of the textures/geometry of a three-dimensional volumetric representation. An encoding computing system can obtain voxel blocks from a three-dimensional volumetric representation of an object. The encoding computing system can encode voxel blocks with a machine-learned voxel encoding model to obtain encoded voxel blocks. The encoding computing system can decode the encoded voxel blocks with a machine-learned voxel decoding model to obtain reconstructed voxel blocks. The encoding computing system can generate a reconstructed mesh representation of the object based at least in part on the one or more reconstructed voxel blocks. The encoding computing system can encode textures associated with the voxel blocks according to an encoding scheme and based at least in part on the reconstructed mesh representation of the object to obtain encoded textures.

Abstract: Apparatus and methods related to applying lighting models to images of objects are provided. An example method includes applying a geometry model to an input image to determine a surface orientation map indicative of a distribution of lighting on an object based on a surface geometry. The method further includes applying an environmental light estimation model to the input image to determine a direction of synthetic lighting to be applied to the input image. The method also includes applying, based on the surface orientation map and the direction of synthetic lighting, a light energy model to determine a quotient image indicative of an amount of light energy to be applied to each pixel of the input image. The method additionally includes enhancing, based on the quotient image, a portion of the input image. One or more neural networks can be trained to perform one or more of the aforementioned aspects.

Abstract: A method for predicting a lower body motion of an avatar is provided. The method includes generating an upper body avatar for a user of a headset, tracking a lower body posture of the user of the headset, retargeting the lower body posture to a lower body model, and merging the lower body model with the upper body avatar to form a full-body avatar for the user of the headset. A system including a memory storing instructions and a processor configured to execute the instructions and cause the system to perform the above method is also provided.

Abstract: A method including generating a first vector based on a first grid and a three-dimensional (3D) position associated with a first implicit representation (IR) of a 3D object, generating at least one second vector based on at least one second grid and an upsampled first grid, decoding the first vector to generate a second IR of the 3D object, decoding the at least one second vector to generate at least one third IR of the 3D object, generating a composite IR of the 3D object based on the second IR of the 3D object and the at least one third IR of the 3D object, and generating a reconstructed volume representing the 3D object based on the composite IR of the 3D object.

Abstract: Techniques of identifying gestures include detecting and classifying inner-wrist muscle motions at a user's wrist using micron-resolution radar sensors. For example, a user of an AR system may wear a band around their wrist. When the user makes a gesture to manipulate a virtual object in the AR system as seen in a head-mounted display (HMD), muscles and ligaments in the user's wrist make small movements on the order of 1-3 mm. The band contains a small radar device that has a transmitter and a number of receivers (e.g., three) of electromagnetic (EM) radiation on a chip (e.g., a Soli chip. This radiation reflects off the wrist muscles and ligaments and is received by the receivers on the chip in the band. The received reflected signal, or signal samples, are then sent to processing circuitry for classification to identify the wrist movement as a gesture.

Abstract: Systems and methods are described for utilizing an image processing system with at least one processing device to perform operations including receiving a plurality of input images of a user, generating a three-dimensional mesh proxy based on a first set of features extracted from the plurality of input images and a second set of features extracted from the plurality of input images. The method may further include generating a neural texture based on a three-dimensional mesh proxy and the plurality of input images, generating a representation of the user including at least a neural texture, and sampling at least one portion of the neural texture from the three-dimensional mesh proxy. In response to providing the at least one sampled portion to a neural renderer, the method may include receiving, from the neural renderer, a synthesized image of the user that is previously not captured by the image processing system.

Abstract: Apparatus and methods related to applying lighting models to images are provided. An example method includes receiving, via a computing device, an image comprising a subject. The method further includes relighting, via a neural network, a foreground of the image to maintain a consistent lighting of the foreground with a target illumination. The relighting is based on a per-pixel light representation indicative of a surface geometry of the foreground. The light representation includes a specular component, and a diffuse component, of surface reflection. The method additionally includes predicting, via the neural network, an output image comprising the subject in the relit foreground. One or more neural networks can be trained to perform one or more of the aforementioned aspects.

Abstract: Embodiments relate to a method for real-time facial animation, and a processing device for real-time facial animation. The method includes providing a dynamic expression model, receiving tracking data corresponding to a facial expression of a user, estimating tracking parameters based on the dynamic expression model and the tracking data, and refining the dynamic expression model based on the tracking data and estimated tracking parameters. The method may further include generating a graphical representation corresponding to the facial expression of the user based on the tracking parameters. Embodiments pertain to a real-time facial animation system.

Abstract: Systems and methods for training models to predict dense correspondences across images such as human images. A model may be trained using synthetic training data created from one or more 3D computer models of a subject. In addition, one or more geodesic distances derived from the surfaces of one or more of the 3D models may be used to generate one or more loss values, which may in turn be used in modifying the model's parameters during training.

Abstract: A head-mounted device (HMD) can be configured to determine a request for recognizing at least one content item included within content framed within a display of the HMD. The HMD can be configured to initiate a head-tracking process that maintains a coordinate system with respect to the content, and a pointer-tracking process that tracks a pointer that is visible together with the content within the display. The HMD can be configured to capture a first image of the content and a second image of the content, the second image including the pointer. The HMD can be configured to map a location of the pointer within the second image to a corresponding image location within the first image, using the coordinate system, and provide the at least one content item from the corresponding image location.

Abstract: Methods, systems, and apparatus including computer programs encoded on a computer storage medium, for generating convex decomposition of objects using neural network models. One of the methods includes receiving an input that depicts an object. The input is processed using a neural network to generate an output that defines a convex representation of the object. The output includes, for each of a plurality of convex elements, respective parameters that define a position of the convex element in the convex representation of the object.

Abstract: Systems and methods for training models to predict dense correspondences across images such as human images. A model may be trained using synthetic training data created from one or more 3D computer models of a subject. In addition, one or more geodesic distances derived from the surfaces of one or more of the 3D models may be used to generate one or more loss values, which may in turn be used in modifying the model's parameters during training.

Abstract: Embodiments relate to a method for real-time facial animation, and a processing device for real-time facial animation. The method includes providing a dynamic expression model, receiving tracking data corresponding to a facial expression of a user, estimating tracking parameters based on the dynamic expression model and the tracking data, and refining the dynamic expression model based on the tracking data and estimated tracking parameters. The method may further include generating a graphical representation corresponding to the facial expression of the user based on the tracking parameters. Embodiments pertain to a real-time facial animation system.

Abstract: Systems and methods for training models to predict dense correspondences across images such as human images. A model may be trained using synthetic training data created from one or more 3D computer models of a subject. In addition, one or more geodesic distances derived from the surfaces of one or more of the 3D models may be used to generate one or more loss values, which may in turn be used in modifying the model's parameters during training.