Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Multi-modal visual concept classification of images via Markov random walk over tags

: Kawanabe, M.; Binder, A.; Müller, C.; Wojcikiewicz, W.


Institute of Electrical and Electronics Engineers -IEEE-:
IEEE Workshop on Applications of Computer Vision, WACV 2011 : Kona, Hawaii, USA, 5 - 7 January 2011; including 2011 IEEE Workshop on Motion and Video Computing (WMVC 2011); part of IEEE Winter Vision Meetings 2011
Piscataway/NJ: IEEE, 2011
ISBN: 978-1-4244-9496-5
ISBN: 978-1-4244-9497-2
Workshop on Applications of Computer Vision (WACV) <2011, Kona/Hawaii>
Workshop on Motion and Video Computing (WMVC) <2011, Kona/Hawaii>
Winter Vision Meetings <2011, Kona/Hawaii>
Conference Paper
Fraunhofer FIRST ()

Automatic annotation of images is a challenging task in computer vision because of "semantic gap" between high-level visual concepts and image appearances. Therefore, user tags attached to images can provide further information to bridge the gap, even though they are partially un-informative and misleading. In this work, we investigate multi-modal visual concept classification based on visual features and user tags via kernel-based classifiers. An issue here is how to construct kernels between sets of tags. We deploy Markov random walks on graphs of key tags to incorporate co-occurrence between them. This procedure acts as a smoothing of tag based features. Our experimental result on the ImageCLEF2010 PhotoAnnotation benchmark shows that our proposed method outperforms the baseline relying solely on visual information and a recently published state-of-the-art approach.