Digital Image Forensic Specialist Uncovers Fake Photos
A team led by University at Albany computer scientist Siwei Lyu has developed a new method for authenticating digital images by analyzing “noise,” the digital equivalent of film grain which is generally invisible to the human eye.
Using statistical and computational analysis, the method effectively measures noise strength across a photo to determine which parts of the photo originated from different sources.
Numerous factors, during and after a photo is taken, introduce noise, such as temperature and thermal conditions, sensor saturation, quantization, compression, and transmission. Since an unaltered image is expected to have uniform noise strength across all pixels, altered photos demonstrate inconsistencies in noise variances.
Whenever a photo is manipulated digitally, the underlying characteristics of the image pixels are disturbed in a way that they become unnatural. Though human eyes may not be able to detect such subtle changes, they can be readily picked up by computer algorithms. The techniques developed by Lyu aim to ensure that significant manipulations can be detected. The new method does not explicitly rely on the knowledge of image format, camera model, or tampering procedure, and has a high level of detection accuracy.
Lyu, the recipient of a highly competitive Faculty Early Career Development Award from the National Science Foundation, developed the new method through his work with the New York State Center for Information Forensics and Assurance.