%Aigaion2 BibTeX export from Idiap Publications %Thursday 21 November 2024 04:51:17 PM @INPROCEEDINGS{gatica03b-conf, author = {Gatica-Perez, Daniel and Sun, Ming-Ting}, projects = {Idiap}, title = {Object Localization in Metric Spaces for Video Linking}, booktitle = {IEEE Workshop on Motion and Video Computing}, year = {2002}, crossref = {gatica03b}, abstract = {While objects often constitute the desired level of access for browsing and retrieval in video databases, an inherent problem for on-line object definition is that of model construction from a few examples. In this paper, we present a probabilistic methodology to localize objects that appear across video segments, based on video structuring, object definition, and localization in the video structure. Localization is formulated as a problem of random sampling in a Metric Mixture Model framework, which allows for the joint modeling of a set of color appearance exemplars and their geometric transformations. To improve the efficiency of the sampling process, candidate configurations are drawn from a prior distribution using importance sampling, and evaluated using Bayes' rule. Experimental results on a database extracted from home videos depicting real objects (with variations of scale and pose) across video shots show the performance of the method.}, pdf = {https://publications.idiap.ch/attachments/reports/2003/rr03-09.pdf}, postscript = {ftp://ftp.idiap.ch/pub/reports/2003/rr03-09.ps.gz}, ipdmembership={vision}, } crossreferenced publications: @TECHREPORT{gatica03b, author = {Gatica-Perez, Daniel and Sun, Ming-Ting}, projects = {Idiap}, title = {Object Localization in Metric Spaces for Video Linking}, type = {Idiap-RR}, number = {Idiap-RR-09-2003}, year = {2003}, institution = {IDIAP}, abstract = {While objects often constitute the desired level of access for browsing and retrieval in video databases, an inherent problem for on-line object definition is that of model construction from a few examples. In this paper, we present a probabilistic methodology to localize objects that appear across video segments, based on video structuring, object definition, and localization in the video structure. Localization is formulated as a problem of random sampling in a Metric Mixture Model framework, which allows for the joint modeling of a set of color appearance exemplars and their geometric transformations. To improve the efficiency of the sampling process, candidate configurations are drawn from a prior distribution using importance sampling, and evaluated using Bayes' rule. Experimental results on a database extracted from home videos depicting real objects (with variations of scale and pose) across video shots show the performance of the method.}, pdf = {https://publications.idiap.ch/attachments/reports/2003/rr03-09.pdf}, postscript = {ftp://ftp.idiap.ch/pub/reports/2003/rr03-09.ps.gz}, ipdmembership={vision}, }