Image Authenticity Verification Using Speech

摘要

Verifying the identity of a person claiming to be represented by a picture by way of providing a string of text (randomly generated or generated by another person seeking verification of same) to be recited by the claimant. The string of text is recited in a video which is received by an intermediary server at a network node, or by a person seeking such verification. Automated processes may be utilized to compare the audio and video received to the picture and string of text sent. Further, comparisons to previously received audio, video, and strings of text, as well as the same available from third parties, may be used to determine fraud attempts. Viewers of the person's profile may also vote on the authenticity of a profile, thereby raising or lowering a certification confidence level, with their votes weighted more heavily towards those who have high confidence levels.

主权项

1. A method of confirming the identity of a person in a picture, comprising the steps of:
receiving at least one picture comprising a version of an individual's likeness via instructions from a first network node associated with a first individual; providing a string of text to said first individual via said first network node; receiving data representative of a video from said first individual; determining, by way of a processor carrying out instructions, whether said video has associated audio representative of said string of text; determining, by way of a processor carrying out instructions, whether said video comprises images of said first individual; providing, via at least a second network node, a certification to at least a second individual that said picture comprising a version of an individual's likeness is that of said first individual.