Abstract: A computing device obtains a video and obtains a target facial image, where the target facial image comprises a target face to be edited. The computing device determines a target facial feature vector from the target facial image to be edited. For each of a plurality of frames in the video, the computing device is further configured to: identify facial regions of individuals depicted in the video, generate candidate facial feature vectors for each of the identified facial regions, compare each of the candidate facial feature vectors to the target facial feature vector, and apply a mask effect to facial regions of corresponding facial feature vectors based on the comparison to generate an edited video.

Abstract: A computing device obtains a live video of a user, where the live video comprises a plurality of image frames of the user. The computing device obtains selection of at least one accessory effect and generates a user interface displaying the live video in a virtual mirror with the at least one accessory effect applied to the user. The computing device monitors for an occurrence of a first target event relating to actions by the user in the live video. In response to detecting an occurrence of the first target event during the live video, the computing device records the live video based on the occurrence of the first target event and detects an occurrence of a second target event relating to actions by the user. The computing device stops recording of the live video responsive to the occurrence of the second target event.

Abstract: A computing device captures a live video of a user, determines a location of a facial region of the user by a facial region analyzer, and determines a finger vector type by a finger vector detector based on a direction in which at least one finger is pointing relative to the facial region of the user. Responsive to detecting a first finger vector type within the facial region involving a single finger, a makeup effects toolbar is displayed in the user interface. Responsive to detecting a second finger vector type involving the single finger, a selection tool for selecting a makeup effect in the makeup effects toolbar is displayed. The computing device obtains a makeup effect based on manipulation by the user of the selection tool. Responsive to detecting a target user action, virtual application of the selected makeup effect is performed on the facial region of the user.

Abstract: A client device obtains a request from a user to retrieve makeup information from a cloud server and obtains a digital image of a facial region of the user. The client device transmits the digital image to initiate a communication session with the cloud server without obtaining login credentials from the user, wherein the cloud server retrieves a makeup session packet by matching the transmitted digital image with one of a plurality of digital images in the cloud server and transmits the makeup session packet comprising a corresponding selection of cosmetic products. The client device receives the makeup session packet and extracts the selection of cosmetic products by the client device and generates a user interface displaying the selection of cosmetic products. The client device performs virtual application of the selection of cosmetic products onto the facial region of the user.

Abstract: A computing device obtains a video, generates a user interface displaying a plurality of facial effects, and obtains selection of one or more facial effects displayed in the user interface. A target facial image is obtained, where the target facial image comprises a target face to be edited. The computing device determines a target facial feature vector from the target facial image to be edited. For each of a plurality of frames in the video, the computing device identifies facial regions of individuals depicted in the video, generates candidate facial feature vectors for each of the identified facial regions, compares each of the candidate facial feature vectors to the target facial feature vector, and applies one or more facial effects to facial regions based on the comparison to generate an edited video.

Abstract: A computing device generates a collection of digital images depicting makeup effects representing makeup trends, analyzes the collection of digital images, and extracts target attributes. The computing device constructs a database of makeup recommendation entries comprising the collection of digital images and extracted attributes and receives a query request from a user comprising an image of the user's face. The computing device queries the database and obtains a first number of makeup recommendations. The computing device merges makeup recommendations among the first number of makeup recommendations to generate a second number of makeup recommendations and displays at least a portion of the second number of makeup recommendations and receiving a selection from the user. The computing device performs virtual application of a makeup effect corresponding to the selection.

Abstract: A makeup evaluation computing device obtains a request from a user to initiate a makeup evaluation session. The makeup evaluation computing device initiates the makeup evaluation session without obtaining login credentials from the user and obtains a digital image of a facial region of the user. The makeup evaluation computing device generates a user interface displaying at least one cosmetic product and obtains from the user a selection of cosmetic products among the one or more displayed cosmetic products. The makeup evaluation computing device performs virtual application of the selected cosmetic products on the facial region of the user and stores the selection of cosmetic products. The makeup evaluation computing device generates a makeup session packet comprising the digital image of the facial region of the user and stored selection of cosmetic products. The makeup evaluation computing device transmits the makeup session packet to a cloud server and terminates the makeup evaluation session.

Abstract: A computing device obtains a live video of a user and obtains selection of at least one makeup effect. The computing device generates a user interface displaying the live video in a virtual mirror with the at least one makeup effect applied to a facial region of the user and monitors for an occurrence of a first target event relating to actions by the user in the live video. In response to detecting an occurrence of the first target event during the live video, the computing device records the live video based on the occurrence of the first target event, detects an occurrence of a second target event relating to actions by the user, and stops recording of the video responsive to occurrence of the second target event.

Abstract: A computing device obtains a digital image depicting an individual and determines lighting conditions of the content in the digital image. The computing device obtains selection of a makeup effect from a user and determines surface properties of the selected makeup effect. The computing device applies a facial alignment technique to the facial region of the individual and defines a region of interest corresponding to the makeup effect. The computing device extracts lighting conditions of the region of interest and adjusts visual characteristics of the makeup effect based on the surface properties of the makeup effect and the lighting conditions of the region of interest. The computing device performs virtual application of the adjusted makeup effect to the region of interest in the digital image.