We have developed algorithms for image similarity computation and image representation that can be used in content based image retrieval systems as well as for image understanding in classification tasks.

• Quicklook2 CBIR system
• Dynamic Spatial Chromatic Histogram
• Prosemantic features

The Quicklook2 CBIR System

Quicklook2 allows the user to query image and multimedia databases with the aid of sample images, or an impromptu sketch and/or textual descriptions, and progressively refine the system’s response by indicating the relevance, or non-relevance of the retrieved items. The major innovation of the system is its relevance feedback mechanism that performs a statistical analysis of both the image and textual feature distributions of the retrieved items the user has judged relevant, or not relevant to identify what features the user has taken into account (and to what extent) in formulating this judgement, and then weigh their influence in the overall evaluation of similarity, as well as in the formulation of a new, single query that better expresses the user’s multimedia information needs.
To see the system at work visit the demo page at this link.

Low level image features: Dynamic Spatial Chromatic Histogram (DSCH)

We introduce an innovative image retrieval strategy, the Dynamic Spatial Chromatic Histogram, which makes it possible to take into account spatial information and dynamic color quantization in a flexible way without greatly adding to computation costs.

High level image features: Prosemantic features

Prosemantic features are obtained through a two-level feature extraction process. A first level of features, related to image structure and color distribution, is extracted from the images, and used as input to a bank of classifiers, each one trained to recognize a given category. Packing together the scores, the features that we call prosemantic are finally obtained. Our experiments show that the use of prosemantic features allows for a more successful and quick retrieval of images with respect to the other features. We have also demonstrate that they can be effectively used also for unsupervised image categorization, that is, for grouping semantically similar images.

@article{ciocca2011halfway-through,
 author = {Ciocca, Gianluigi and Cusano, Claudio and Santini, Simone and Schettini, Raimondo},
 year = {2011},
 pages = {4943-4958},
 title = {Halfway through the semantic gap: Prosemantic features for image retrieval},
 volume = {181},
 number = {22},
 publisher = {Elsevier},
 journal = {Information Sciences},
 url = {http://www.sciencedirect.com/science/article/pii/S002002551100332X},
 pdf = {/download/ciocca2011halfway-through.pdf},
 doi = {10.1016/j.ins.2011.06.025},
 issn = {0020-0255}}

@article{ciocca2006supervised-unsupervised,
 author = {Ciocca, Gianluigi and Schettini, Raimondo},
 year = {2006},
 pages = {630-638},
 title = {Supervised and unsupervised classification post-processing for visual video summaries},
 volume = {52},
 number = {2},
 publisher = {IEEE},
 journal = {IEEE Transactions on Consumer Electronics},
 url = {http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=1649689},
 pdf = {/download/ciocca2006supervised-unsupervised.pdf},
 doi = {10.1109/TCE.2006.1649689},
 issn = {0098-3063}}

@inproceedings{ciocca2003image-retrieval,
 author = {Ciocca, Gianluigi and Schettini, Raimondo},
 year = {2003},
 pages = {1829-1837},
 title = {Image retrieval using dynamic spatial chromatic histograms},
 volume = {5150},
 publisher = {{SPIE}},
 booktitle = {Visual Communications and Image Processing 2003},
 url = {http://link.aip.org/link/?PSI/5150/1829/1},
 doi = {10.1117/12.503344}}

@article{ciocca2001quicklook2-integrated,
 author = {Ciocca, Gianluigi and Gagliardi, Isabella and Schettini, Raimondo},
 year = {2001},
 pages = {81-103},
 title = {Quicklook2: An Integrated Multimedia System},
 volume = {12},
 number = {1},
 publisher = {Elsevier},
 journal = {Journal of Visual Languages and Computing},
 url = {http://www.sciencedirect.com/science/article/pii/S1045926X00901885},
 pdf = {/download/ciocca2001quicklook2-integrated.pdf},
 doi = {10.1006/jvlc.2000.0188},
 issn = {1045-926X}}