11 February 2011 Image retrieval considering people co-occurrence relations using relevance feedback
Author Affiliations +
The recent popularity of digital cameras allows us to take a large number of images. There is an increasing need for efficiently and accurately retrieving images containing a specific person from such image collections. While only the visual features of the specific person are used in many query-by-example retrieval methods, we focus on the fact that some people such as family or friends are more likely to appear in the same images than others and use visual features of not only the queried person but also people who have strong co-occurrence relations with the queried person to improve the retrieval performance. The relevance feedback is used to learn who co-occur with the queried person in the same images, their faces, and the strength of their co-occurrence relations. For 116 images collected from 6 persons, after five feedback iterations, the recall rate of 53% was obtained by considering the co-occurrence relations among people, as against 34% when using only features of the queried person.
© (2011) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Kazuya Shimizu, Kazuya Shimizu, Naoko Nitta, Naoko Nitta, Noboru Babaguchi, Noboru Babaguchi, } "Image retrieval considering people co-occurrence relations using relevance feedback", Proc. SPIE 7881, Multimedia on Mobile Devices 2011; and Multimedia Content Access: Algorithms and Systems V, 788117 (11 February 2011); doi: 10.1117/12.872125; https://doi.org/10.1117/12.872125


Fermat theorem and elliptic color histogram features
Proceedings of SPIE (January 12 2003)
UCDL:a language for audio-visual content description
Proceedings of SPIE (August 23 1999)
Content-based video retrieval via motion trajectories
Proceedings of SPIE (August 18 1998)

Back to Top