Automatic image representation and clustering on mobile devices
Journal of Mobile Multimedia
Braving the semantic gap: mapping visual concepts from images and videos
ICDM'04 Proceedings of the 4th international conference on Advances in Data Mining: applications in Image Mining, Medicine and Biotechnology, Management and Environmental Control, and Telecommunications
Unsupervised clustering in personal photo collections
AMR'08 Proceedings of the 6th international conference on Adaptive Multimedia Retrieval: identifying, Summarizing, and Recommending Image and Music
| Hi-index | 0.00 |
Content-based image retrieval techniques have been underintensively research, mainly on on extracting effectivelow level visual features for indexing and enabling fastquery of individual images by feature matching over theindexing structure. In this paper we propose to extend thecontent-based approach towards the problem of multimediacollection profiling and comparison. Our method is to carryout visual feature clustering using self-organised maps, andthen apply distance measures on the generated feature mapsto evaluate their similarity. A modified Hausdorff distance isdefined over the feature maps and further verified in an experimentusing four image collections. Some preliminary resultsare presented with a comparison of different distancemeasures obtained from profiles generated by two featureschemes.