Machine learning based makeup look recommendation and routine application method

A system comprising:

• one or more processors coupled to memory, the memory loaded with computer instructions that, when executed on the one or more processors, implement actions comprising:
• receiving user-specific behavioral history data, one or more user-input assessments specifying aesthetic preferences of a user, and data trends from a plurality of users within a given demographic;
• generating, by makeup recommendation logic comprising a trained machine learning model, one or more makeup look recommendations based at least in part on the user-specific behavioral history data, the one or more user-input assessments, and the data trends;
• selecting, from a database of makeup looks and based on the one or more makeup look recommendations, a catalogue of makeup looks that presents respective makeup routines, each respective makeup routine executable to achieve a corresponding makeup look in the catalogue;
• receiving a selection of at least one makeup routine in the presented respective makeup routines;
• and
• providing step-by-step instructions for executing a selected makeup routine, the step-by-step instructions presented in a graphical user interface that displays, for a particular step in the step-by-step instructions, a graphical simulation of how execution of the particular step is rendered as applied to a graphical representation of the user, and updating the graphical simulation as the user advances from the particular step to a subsequent step in the step-by-step instructions.

A non-transitory computer readable storage medium impressed with computer program instructions that, when executed on a processor, implement a method comprising:

• receiving user-specific behavioral history data, one or more user-input assessments specifying aesthetic preferences of a user, and data trends from a plurality of users within a given demographic;
• generating, by makeup recommendation logic, one or more makeup look recommendations in response to processing, as input, the user-specific behavioral history data, the one or more user-input assessments, and the data trends, wherein the makeup recommendation logic weights respective descriptive features, combined into a feature vector, by respective feature importances learned by a machine learning model;
• selecting, from a database of makeup looks and based on the one or more makeup look recommendations, a catalogue of makeup looks that presents respective makeup routines, each respective makeup routine executable to achieve a corresponding makeup look in the catalogue;
• receiving a selection of at least one makeup routine in the presented respective makeup routines;
• and
• providing step-by-step instructions for executing a selected makeup routine, the step-by-step instructions presented in a graphical user interface that displays, for a particular step in the step-by-step instructions, a graphical simulation of how execution of the particular step is rendered.