|        Evaluation of Localized Semantics in Images    |
In this project, we create a new data set of 1014 images with manual segmentations and semantic labels for each segment, and present a methodology for using this kind of data for recognition evaluation. The evaluation methodology establishes protocols for mapping machine segmentation to human segmentation, scoring matches at different levels of specificity, and taking synonyms, sense ambiguity and multiple labels into accounted. Based on these protocols, we develop two evaluation approaches for measuring the range and the frequency of semantics that an algorithm can recognize correctly.
The images and segmentations are from the UCB segmentation benchmark database (Martin et al., 2001). The database is extended by manually labeling each segment with its most specific semantic concept in WordNet (See  for details). The data set is available HERE. The data format is described HERE and the labeling rules can be found HERE.
There are two software tools provided here,
The project is a collaboration between the Computer Science Department and GE Global Research.
|Questions or Comments
If any questions or comments, please email Kobus Barnard at kobus AT sista DOT arizona DOT edu.
Kobus Barnard, Quanfu Fan, Ranjini Swaminathan, Anthony Hoogs, Roderic Collins, Pascale Rondot, John Kaufhold, "Evaluation of localized semantics: data, methodology, and experiments," International Journal of Computer Vision, Vol. 77, pp 199-217, 2008.    [ PDF]
 Kobus Barnard, Quanfu Fan, Ranjini Swaminathan, Anthony Hoogs, Roderic Collins, Pascale Rondot, John Kaufhold, "Evaluation of localized semantics: data, methodology, and experiments", International Journal of Computer Vision (to appear) PDF.