Techniques are generally described for user authentication. Example techniques may include providing a data set including audio data and image data, wherein the audio data includes voice recordings of multiple people, wherein the image data includes at least a facial image of at least one of the multiple people, receiving a response to the data set from a user device, and determining whether the received response corresponds to at least a part of content of the voice recording of the one of the multiple people whose facial image is included in the image data.