CBIAProto3 is a Content-Based Image Annotation engine initially based on the method proposed in the article:

@conference{pan2005gcap,
  title={Gcap: Graph-based automatic image captioning},
  author={Pan, J.Y. and Yang, H.J. and Faloutsos, C. and Duygulu, P.},
  booktitle={Computer Vision and Pattern Recognition Workshop, 2004. CVPRW'04. Conference on},
  pages={146},
  year={2005},
  organization={IEEE}
}

and used in some of my articles like:

@conference{jamesCIAM2009,
  title={Towards Semantic Image Annotation With Keyword Disambiguation Using Semantic And Visual Knowledge},
  author={N. James and C. Hudelot},
  booktitle={Proceedings of the IJCAI workshop on Cross-media Information Access And Mining, CIAM 2009, Pasadena, USA},
  pages={35--40},  
  year={2009},
}

The software use a specific image index model named IIDF (Image Index and Data File) and has been developed for exploring image annotation using multi-modal learning data.

Online demonstrator here (with a noisy vocabulary and a small learning dataset due to limited computing resources, a better model will be uploaded soon... ).

A firefox plugin (compatible with firefox 7): trevizIt.xpi
It adds an option in the contextual menu of images included in a web page.

WNXplorer is an interactive and graph-based navigation software for the WordNet database.

 

DatasetExplorer is a Java application for exploring large collections of annotated images. The program is based on the Galatee library.

Howto use it to visualize the PascalVOC data. Now visualizing the PascalVOC data is integrated to DatasetExplorer.

 

LabelMeWorkbench is a toolbox for (1) create indices of the annotations of the LabelMe dataset, (2) browse the dataset, and (3) perform queries on the indices.

[  LabelMeWorkbench-0.45-2010.07.08-SNAPSHOT.tar.bz2 ]

WNLXplorer is a WordNet visualization application and toolbox based on Jung2 and JWI and Lucene.
WordNet 3.0 is entirely indexed in a Lucene index, this allows to add payloads to synsets or update the synsets (adding co-occurrence information or statistical information, defining new semantic relations etc.), and to add new synsets.

See this wiki page for the full list of my projects.