Comparing discriminating transformations and SVM for learning during multimedia retrieval
MULTIMEDIA '01 Proceedings of the ninth ACM international conference on Multimedia
Collaterally cued labelling framework underpinning semantic-level visual content descriptor
VISUAL'07 Proceedings of the 9th international conference on Advances in visual information systems
Hi-index | 0.00 |
In this paper feature, selection and representation techniques in CBIR systems are reviewed and interpreted in a unified feature representation paradigm. We revise our previously proposed water-filling edge features with newly proposed primitives and present them using this unified feature formation paradigm. Experiments and comparisons are performed to illustrate the characteristics of the new features. Also proposed is sub-image feature extraction for regional matching. Relevance feedback as an on-line learning mechanism is adopted for feature and tile selection and weighting during the retrieval.