In visual object retrieval, there are always queries which cannot be processed successfully. In these cases, it is desirable to extract imprecise symbolic information. We present an architecture that extracts symbolic descriptors of the objects shown in the query image. The method is based on a combination of numeric feature extraction and classification. We describe some examples of descriptors and present first experimental results. © Springer-Verlag 2003.
CITATION STYLE
Wickel, J., Alvarado, P., & Kraiss, K. F. (2003). Extracting symbolic descriptors for interactive object retrieval. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2749, 290–297. https://doi.org/10.1007/3-540-45103-x_40
Mendeley helps you to discover research relevant for your work.