DocumentCode :
3427250
Title :
From Point to Set: Extend the Learning of Distance Metrics
Author :
Pengfei Zhu ; Lei Zhang ; Wangmeng Zuo ; Zhang, Dejing
Author_Institution :
Hong Kong Polytech. Univ., Hong Kong, China
fYear :
2013
fDate :
1-8 Dec. 2013
Firstpage :
2664
Lastpage :
2671
Abstract :
Most of the current metric learning methods are proposed for point-to-point distance (PPD) based classification. In many computer vision tasks, however, we need to measure the point-to-set distance (PSD) and even set-to-set distance (SSD) for classification. In this paper, we extend the PPD based Mahalanobis distance metric learning to PSD and SSD based ones, namely point-to-set distance metric learning (PSDML) and set-to-set distance metric learning (SSDML), and solve them under a unified optimization framework. First, we generate positive and negative sample pairs by computing the PSD and SSD between training samples. Then, we characterize each sample pair by its covariance matrix, and propose a covariance kernel based discriminative function. Finally, we tackle the PSDML and SSDML problems by using standard support vector machine solvers, making the metric learning very efficient for multiclass visual classification tasks. Experiments on gender classification, digit recognition, object categorization and face recognition show that the proposed metric learning methods can effectively enhance the performance of PSD and SSD based classification.
Keywords :
covariance matrices; image classification; learning (artificial intelligence); PPD based Mahalanobis distance metric learning; PSD; PSDML; SSD; SSDML; covariance matrix; digit recognition; face recognition; gender classification; multiclass visual classification; object categorization; point-to-point distance; point-to-set distance; point-to-set distance metric learning; set-to-set distance; set-to-set distance metric learning; support vector machine solvers; Covariance matrices; Face recognition; Learning systems; Measurement; Support vector machines; Training; Visualization;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer Vision (ICCV), 2013 IEEE International Conference on
Conference_Location :
Sydney, NSW
ISSN :
1550-5499
Type :
conf
DOI :
10.1109/ICCV.2013.331
Filename :
6751442
Link To Document :
بازگشت