Similarity comparisons for interactive fine-grained categorization
Research output: Contribution to journal › Conference article › Research › peer-review
Standard
Similarity comparisons for interactive fine-grained categorization. / Wah, Catherine; Horn, Grant Van; Branson, Steve; Maji, Subhransu; Perona, Pietro; Belongie, Serge.
In: Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 24.09.2014, p. 859-866.Research output: Contribution to journal › Conference article › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - GEN
T1 - Similarity comparisons for interactive fine-grained categorization
AU - Wah, Catherine
AU - Horn, Grant Van
AU - Branson, Steve
AU - Maji, Subhransu
AU - Perona, Pietro
AU - Belongie, Serge
N1 - Publisher Copyright: © 2014 IEEE.
PY - 2014/9/24
Y1 - 2014/9/24
N2 - Current human-in-the-loop fine-grained visual categorization systems depend on a predefined vocabulary of attributes and parts, usually determined by experts. In this work, we move away from that expert-driven and attribute-centric paradigm and present a novel interactive classification system that incorporates computer vision and perceptual similarity metrics in a unified framework. At test time, users are asked to judge relative similarity between a query image and various sets of images, these general queries do not require expert-defined terminology and are applicable to other domains and basic-level categories, enabling a flexible, efficient, and scalable system for fine-grained categorization with humans in the loop. Our system outperforms existing state-of-the-art systems for relevance feedback-based image retrieval as well as interactive classification, resulting in a reduction of up to 43% in the average number of questions needed to correctly classify an image.
AB - Current human-in-the-loop fine-grained visual categorization systems depend on a predefined vocabulary of attributes and parts, usually determined by experts. In this work, we move away from that expert-driven and attribute-centric paradigm and present a novel interactive classification system that incorporates computer vision and perceptual similarity metrics in a unified framework. At test time, users are asked to judge relative similarity between a query image and various sets of images, these general queries do not require expert-defined terminology and are applicable to other domains and basic-level categories, enabling a flexible, efficient, and scalable system for fine-grained categorization with humans in the loop. Our system outperforms existing state-of-the-art systems for relevance feedback-based image retrieval as well as interactive classification, resulting in a reduction of up to 43% in the average number of questions needed to correctly classify an image.
UR - http://www.scopus.com/inward/record.url?scp=84911368243&partnerID=8YFLogxK
U2 - 10.1109/CVPR.2014.115
DO - 10.1109/CVPR.2014.115
M3 - Conference article
AN - SCOPUS:84911368243
SP - 859
EP - 866
JO - I E E E Conference on Computer Vision and Pattern Recognition. Proceedings
JF - I E E E Conference on Computer Vision and Pattern Recognition. Proceedings
SN - 1063-6919
T2 - 27th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2014
Y2 - 23 June 2014 through 28 June 2014
ER -
ID: 302044146