Semantic visual templates: linking visual features to semantics
- 27 November 2002
- proceedings article
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 3, 531-535
- https://doi.org/10.1109/icip.1998.727321
Abstract
The rapid growth of visual data over the last few years has lead to many schemes for retrieving such data. With content-based systems today, there exists a significant gap between the user's information needs and what the systems can deliver. We propose to bridge this gap, by introducing the novel idea of Semantic Visual Templates (SVT). Each template represents a personalized view of concepts (e.g. slalom, meetings, sunsets etc.), The SVT is represented us- ing a set of successful queries, which are generated by a two-way interaction between the user and the system. We have developed algorithms that interact with the user and converge upon a small set of exemplar queries that maxi- mize recall. SVT's emphasize intuitive models that allow for easy manipulation and queries to be composited. The resulting system performs well, for example with small num- ber of queries in the "sunset" template, we are able to achieve 50% recall and 24% precision over a large unan- notated database.Keywords
This publication has 6 references indexed in Scilit:
- Target testing and the PicHunter Bayesian multimedia retrieval systemPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- On image classification: city vs. landscapePublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- Virage video enginePublished by SPIE-Intl Soc Optical Eng ,1997
- Visual image retrieval by elastic matching of user sketchesPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1997
- VideoQPublished by Association for Computing Machinery (ACM) ,1997
- VisualSEEkPublished by Association for Computing Machinery (ACM) ,1996