Fuente: WIPO "tomato" An example process includes: detecting first image data that represents a first scene; detecting audio data, wherein: the audio data includes a speech input; and the speech input includes a user request to perform a task; and in response to detecting the audio data and after a determination that the first scene includes a user who provides the speech input: in accordance with a determination that an identification score satisfies a first threshold score, wherein the identification score represents a confidence that a user is a registered user of the electronic device, initiating the task, wherein the task is personalized for the registered user; and in accordance with a determination that the identification score does not satisfy the first threshold score and in accordance with a determination that a set of identification score adjustment criteria is satisfied: initiating an action; and adjusting the identification score based on the initiated action.