US Patent 12073839 Authentication by speech at a machine

In some implementations, a front-end device may receive a physical identifier associated with the user. Accordingly, the front-end device may select a plurality of images, where each image corresponds to a unique integer of integers zero through nine. The front-end device may show, on a display, the plurality of images and receive audio that includes a sequence of words that describe a subset of the plurality of images. Accordingly, the front-end device may map the sequence of words to the subset of the plurality of images and determine a first sequence of numbers corresponding to the subset of the plurality of images. Therefore, the front-end device may authenticate the user based on the first sequence of numbers matching a second sequence of numbers associated with the user.