Abstract: Novel tools and techniques are provided for implementing image authentication for authenticating persons or items. In various embodiments, a computing system might receive an image (or video stream) from an optical sensor device, might extract image regions from each received image (or each image frame from each image stream), might analyze each image region to identify one or more spatial relationships amongst pixels (and/or groups of pixels) in each image region, might compare each identified spatial relationship amongst (groups of) pixels in each image region with a plurality of spatial relationships amongst (groups of) pixels that are characteristic of particular image artifacts (whether known and/or machine-learned), and might generate authenticity values and/or results for the image based at least in part on results of the analysis and the comparison. The computing system might also analyze a plurality of authentic and inauthentic images to identify distinctions between authentic and inauthentic images.