DocumentCode :
3294162
Title :
Social Image Tagging by Mining Sparse Tag Patterns from Auxiliary Data
Author :
Jie Lin ; Junsong Yuan ; Ling-Yu Duan ; Siwei Luo ; Wen Gao
Author_Institution :
Sch. of Comput. & Inf. Technol., Beijing Jiaotong Univ., Beijing, China
fYear :
2012
fDate :
9-13 July 2012
Firstpage :
7
Lastpage :
12
Abstract :
User-given tags associated with social images from photosharing websites (e.g., Flickr) are valuable auxiliary resources for the image tagging task. However, social images often suffer from noisy and incomplete tags, heavily degrading the effectiveness of previous image tagging approaches. To alleviate the problem, we introduce a Sparse Tag Patterns (STP) model to discover noiseless and complementary cooccurrence tag patterns from large scale user contributed tags among auxiliary web data. To fulfill the compactness and discriminability, we formulate the STP model as a problem of minimizing quadratic loss function regularized by bi-layer ℓ1 norm. We treat the learned STP as a universal knowledge base and verify its superiority within a data-driven image tagging framework. Experimental results over 1 million auxiliary data demonstrate superior performance of the proposed method compared to the state-of-the-art.
Keywords :
data mining; image retrieval; social networking (online); STP model; auxiliary Web data; data-driven image tagging framework; incomplete tags; large scale user contributed tag; noisy tags; photosharing websites; quadratic loss function; social image tagging; sparse tag pattern mining; user-given tags; Educational institutions; Encoding; Image color analysis; Noise measurement; Optimization; Tagging; Visualization; Auxiliary Data; CBIR; Social Image Tagging; Sparse Tag Pattern;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Multimedia and Expo (ICME), 2012 IEEE International Conference on
Conference_Location :
Melbourne, VIC
ISSN :
1945-7871
Print_ISBN :
978-1-4673-1659-0
Type :
conf
DOI :
10.1109/ICME.2012.170
Filename :
6298366
Link To Document :
بازگشت