Digital watermarking and soft biometrics

Angela D'Angelo - PhD student in Multimedia Communications
Multimedia Communications

Date: -
Location: Eurecom

The first part of the talk will focus on the characterization and quality evaluation of geometric distortions in images with application to digital watermarking. In fact, a first step to solve the problems with geometric attacks in digital watermarking is the characterization of the class of perceptually admissible distortions. This requires the development of models to treat the distortions from a mathematical point of view, and the definition of an objective quality metric for the perceptual evaluation of geometrically distorted images. In the second part of the talk we will see the application of color based soft biometry for hooligans detection. Biometric systems based on human traits like fingerprinting, face, iris, etc., even widely explored in literaure, cannot be used in scenarios like video surveillance of crowded enviroments. A new class of biometry, usually referred as soft biometry, including individual’s height, weight, skin color, hair color, clothes color, motion trajectory, etc., can be, instead, easily extracted from cameras in a large network to provide some information about the users. Among soft biometric signatures, we focus our attention on color. Based on the idea that sport fans usually wear clothes of the same color of the team they support, we provide a video analysis framework for hooligans detection.