DocumentCode :
3427060
Title :
Visual Reranking through Weakly Supervised Multi-graph Learning
Author :
Cheng Deng ; Rongrong Ji ; Wei Liu ; Dacheng Tao ; Xinbo Gao
Author_Institution :
Xidian Univ., Xi´an, China
fYear :
2013
fDate :
1-8 Dec. 2013
Firstpage :
2600
Lastpage :
2607
Abstract :
Visual reranking has been widely deployed to refine the quality of conventional content-based image retrieval engines. The current trend lies in employing a crowd of retrieved results stemming from multiple feature modalities to boost the overall performance of visual reranking. However, a major challenge pertaining to current reranking methods is how to take full advantage of the complementary property of distinct feature modalities. Given a query image and one feature modality, a regular visual reranking framework treats the top-ranked images as pseudo positive instances which are inevitably noisy, difficult to reveal this complementary property, and thus lead to inferior ranking performance. This paper proposes a novel image reranking approach by introducing a Co-Regularized Multi-Graph Learning (Co-RMGL) framework, in which the intra-graph and inter-graph constraints are simultaneously imposed to encode affinities in a single graph and consistency across different graphs. Moreover, weakly supervised learning driven by image attributes is performed to denoise the pseudo-labeled instances, thereby highlighting the unique strength of individual feature modality. Meanwhile, such learning can yield a few anchors in graphs that vitally enable the alignment and fusion of multiple graphs. As a result, an edge weight matrix learned from the fused graph automatically gives the ordering to the initially retrieved results. We evaluate our approach on four benchmark image retrieval datasets, demonstrating a significant performance gain over the state-of-the-arts.
Keywords :
content-based retrieval; feature extraction; image denoising; image fusion; learning (artificial intelligence); Co-RMGL framework; co-regularized multigraph learning; content-based image retrieval engines; distinct feature modalities; edge weight matrix; image reranking approach; image retrieval datasets; inter-graph constraints; intra-graph constraints; multiple feature modalities; multiple graph alignment; multiple graph fusion; pseudo-labeled instances; query image; visual reranking; weakly supervised multigraph learning; Image edge detection; Labeling; Noise measurement; Semantics; Supervised learning; Vectors; Visualization; Visual reranking; attribute; graph anchor; multi-graph learning; weakly-supervised learning;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer Vision (ICCV), 2013 IEEE International Conference on
Conference_Location :
Sydney, VIC
ISSN :
1550-5499
Type :
conf
DOI :
10.1109/ICCV.2013.323
Filename :
6751434
Link To Document :
بازگشت