Abstract: A method for training a machine learning model using information pertaining to a human face, the method includes generating training data for the machine learning model. Generating the training data includes generating a training input, the training input including information representing 2D images of human faces corresponding to a beauty target, and generating a target output for the training input. The target output identifies, for each of the 2D images of human faces corresponding to the beauty target, information identifying one or more facial features represented in the respective 2D image of human faces corresponding to the beauty target. The method further includes providing the training data to train the machine learning model on (i) a set of training inputs including the training input, and (ii) a set of target outputs including the target output.

Abstract: A method for training a machine learning model using information pertaining to an eye area of a human face, the method includes generating training data for the machine learning model. Generating the training data includes generating a first training input, the first training input including information representing 2D images of human faces, and generating a first target output for the training input. The first target output identifies, for each of the 2D images of human faces, information identifying an eye shape category of multiple eye shape categories, the eye shape category corresponding to a shape of an eye represented in the respective 2D image. The method further includes providing the training data to train the machine learning model on (i) a set of training inputs including the first training input, and (ii) a set of target outputs including the first target output.

Abstract: A method includes receiving, by a processing device and from a client device, a first video segment of a video stream. The method further includes generating, using the first video segment, a three-dimensional (3D) model of a human face of a subject. The method further includes modifying the first video segment to include i) a first overlay of the 3D model identifying one or more landmarks and ii) first visual indicia positioned in the first video segment based on the one or more landmarks of the 3D model and related to a first operation of a plurality of operations of an application of artificial lash extensions. The method further includes providing the first modified video segment for presentation at the client device, the first modified video segment including the first visual indicia related to the first operation of the application of artificial lash extensions.

Abstract: A first computer vision operation is performed on a first video segment of a video stream to track points corresponding to an underside of natural eyelashes of an eye area of a subject. The first video segment is modified to include a first overlay of a first model of a first artificial lash extension positioned at least at a first point of the points corresponding to the underside of the natural eyelashes. The first modified video segment is provided for presentation at a client device associated with the subject.

Abstract: Two-dimensional (2D) image data corresponding to a 2D image of an eye area of a subject is received. Three-dimensional (3D) data representing the eye area of the subject is identified based on the 2D image data. A first landmark representing a first facial feature of the eye area, a second landmark representing a second facial feature of the eye area, and a relationship between the first landmark and the second landmark is determined based on the 3D data representing the eye area of the subject. A subset of the plurality of false eyelashes comprising a subset of a plurality of artificial lash extensions designed for application at an underside of natural eyelashes of the subject are identified, among a plurality of false eyelashes, based on the relationship between the first landmark and the second landmark.

Abstract: A method includes performing, by a processing device, an image processing operation on image data corresponding to an image representing a human face to determine a textual identifier that describes at least one facial feature of the human face. The method further includes generating a prompt for a generative machine learning model. The prompt includes (i) information corresponding to the textual identifier and (ii) beauty product information indicative of multiple beauty products. The method further includes obtaining, from the generative machine learning model and based on the prompt, an output identifying a set of beauty products related to the at least one facial feature of the human face.

Abstract: A method includes receiving, by a processing device and from a client device, a first video segment of a video stream. The method further includes generating, using the first video segment, a three-dimensional (3D) model of a human face of a subject. The method further includes modifying the first video segment to include i) a first overlay of the 3D model identifying one or more landmarks and ii) first visual indicia positioned in the first video segment based on the one or more landmarks of the 3D model and related to a first operation of a plurality of operations of an application of artificial lash extensions. The method further includes providing the first modified video segment for presentation at the client device, the first modified video segment including the first visual indicia related to the first operation of the application of artificial lash extensions.

Abstract: A first computer vision operation is performed on a first video segment of a video stream to track points corresponding to an underside of natural eyelashes of an eye area of a subject. The first video segment is modified to include a first overlay of a first model of a first artificial lash extension positioned at least at a first point of the points corresponding to the underside of the natural eyelashes. The first modified video segment is provided for presentation at a client device associated with the subject.

Abstract: A method for training a machine learning model using information pertaining to an eye area of a human face, the method includes generating training data for the machine learning model. Generating the training data includes generating a first training input, the first training input including information representing 2D images of human faces, and generating a first target output for the training input. The first target output identifies, for each of the 2D images of human faces, information identifying an eye shape category of multiple eye shape categories, the eye shape category corresponding to a shape of an eye represented in the respective 2D image. The method further includes providing the training data to train the machine learning model on (i) a set of training inputs including the first training input, and (ii) a set of target outputs including the first target output.

Abstract: A method includes determining a textual identifier that describes at least one facial feature of a human face based on two-dimensional image data representing the human face. The method further includes generating a prompt for a generative machine learning model. The prompt includes information corresponding to the textual identifier that describes the at least one facial feature of the human face. The method further includes obtaining, from the generative machine learning model and based on the prompt, an output indicative of a set of false eyelashes that suit the human face.

Abstract: A method for training a machine learning model using information pertaining to a human face, the method includes generating training data for the machine learning model. Generating the training data includes generating a training input, the training input including information representing 2D images of human faces corresponding to a beauty target, and generating a target output for the training input. The target output identifies, for each of the 2D images of human faces corresponding to the beauty target, information identifying one or more facial features represented in the respective 2D image of human faces corresponding to the beauty target. The method further includes providing the training data to train the machine learning model on (i) a set of training inputs including the training input, and (ii) a set of target outputs including the target output.

Abstract: A first two-dimensional (2D) image data corresponding to a first 2D image of a human face of a subject is received. First metrics corresponding to a first target model of a target human face are identified among multiple target models. Second metrics that represent one or more facial features of the human face of the subject are generated based on the first 2D image data. Variation information identifying one or more differences between the first metrics corresponding to the first target model of the target human face and the second metrics that represent the one or more facial features of the human face of the subject is determined. A notification reflecting the variation information is provided to a client device.

Abstract: Two-dimensional (2D) image data corresponding to a 2D image of an eye area of a subject is received. A three-dimensional (3D) model of the eye area is generated using the 2D image data. A relationship between a first landmark of the 3D model and a second landmark of the 3D model is determined. The first landmark represents a first facial feature of the eye area, and the second landmark represents a second facial feature of the eye area. A subset of a plurality of false eyelashes is identified among a plurality of false eyelashes based on the relationship between the first landmark and the second landmark of the 3D model.

Abstract: A method includes receiving 2D image data corresponding to a 2D image of a human face. The method further includes determining a textual identifier that describes a facial feature of the human face based on the 2D image data. The method further includes providing, to a generative machine learning model, a first prompt including information identifying the textual identifier that describes the facial feature of the human face. The method further includes obtaining, from the generative machine learning model, a first output identifying, among a plurality of beauty products, a subset of the plurality of beauty products, the subset of the plurality of beauty products related to the facial feature of the human face.

Abstract: A method includes receiving, by a processing device and from a client device, a first video segment of a video stream, wherein the first video segment represents an eye area of a subject. The method further includes performing a first computer vision operation on the first video segment to track a plurality of points corresponding to the eye area of the subject. The method further includes modifying the first video segment to include an overlay of a first 3D model of false eyelashes proximate the tracked points corresponding to the eye area of the subject. The method further includes providing the first modified video segment for presentation at the client device.

Abstract: A method includes receiving 2D image data corresponding to a 2D image of a human face of a subject. The method further includes determining a textual identifier that describes a facial feature of the human face represented by the 2D image. The method further includes providing, to a first generative machine learning model, a first prompt including information identifying the textual identifier that describes the facial feature of the human face. The method further includes obtaining, from the first generative machine learning model, a first output identifying, among a plurality of sets of false eyelashes, a set of false eyelashes selected to suit the human face of the subject based on the facial feature.

Abstract: A system to process images includes a light source configured to emit a first illumination pattern onto one or more first portions of a scene. The system also includes an image sensor configured to capture light reflected from the scene in response to the emitted first illumination pattern. The system also includes an optimizer configured to perform raytracing of the light reflected from the scene. The system further includes a processor operatively coupled to the optimizer. The processor is configured to determine a parameter of a surface of the scene based on the raytracing, cause the light source to emit a second illumination pattern onto one or more second portions of the scene based at least in part on the parameter of the surface, and refine the parameter of the surface of the scene based on additional raytracing performed on reflected light from the second illumination pattern.

Abstract: Approaches in accordance with various embodiments provide for the presentation of augmented reality (AR) content with respect to optically challenging surfaces. Such surfaces can be difficult to locate using conventional optical-based approaches that rely on visible features. Embodiments can utilize the fact that horizontal surfaces can be located relatively easily, and can determine intersections or boundaries of those horizontal surfaces that likely indicate the presence of another surface, such as a vertical wall. This boundary can be determined automatically, through user input, or using a combination of such approaches. Once such an intersection is located, a virtual plane can be determined whose relative location to a device displaying AR content can be tracked and used as a reference for displaying AR content.

Abstract: Approaches in accordance with various embodiments provide for the presentation of augmented reality (AR) content with respect to optically challenging surfaces. Such surfaces can be difficult to locate using conventional optical-based approaches that rely on visible features. Embodiments can utilize the fact that horizontal surfaces can be located relatively easily, and can determine intersections or boundaries of those horizontal surfaces that likely indicate the presence of another surface, such as a vertical wall. This boundary can be determined automatically, through user input, or using a combination of such approaches. Once such an intersection is located, a virtual plane can be determined whose relative location to a device displaying AR content can be tracked and used as a reference for displaying AR content.

Abstract: A system to process images includes a light source configured to emit a first illumination pattern onto one or more first portions of a scene. The system also includes an image sensor configured to capture light reflected from the scene in response to the emitted first illumination pattern. The system also includes an optimizer configured to perform raytracing of the light reflected from the scene. The system further includes a processor operatively coupled to the optimizer. The processor is configured to determine a parameter of a surface of the scene based on the raytracing, cause the light source to emit a second illumination pattern onto one or more second portions of the scene based at least in part on the parameter of the surface, and refine the parameter of the surface of the scene based on additional raytracing performed on reflected light from the second illumination pattern.