Abstract

In this paper we present an overview of a software platform that has been developed within the aceMedia project, termed the aceToolbox, that provides global and local lowlevel feature extraction from audio-visual content. The toolbox is based on the MPEG-7 eXperimental Model (XM), with extensions to provide descriptor extraction from arbitrarily shaped image segments, thereby supporting local descriptors reflecting real image content. We describe the architecture of the toolbox as well as providing an overview of the descriptors supported to date. We also briefly describe the segmentation algorithm provided. We then demonstrate the usefulness of the toolbox in the context of two different content processing scenarios: similarity-based retrieval in large collections and scene-level classification of still images.