Transferring knowledge fragments for learning distance metric from a heterogeneous domain

• Published in: IEEE Transactions on Pattern Analysis and Machine Intelligence
• Main Authors: Luo, Yong, Wen, Yonggang, Liu, Tongliang, Tao, Dacheng
• Other Authors: School of Computer Science and Engineering
• Format: Article in Journal/Newspaper
• Language: English
• Published: 2019
• Subjects: Statistics - Machine Learning; Computer Science - Computer Vision and Pattern Recognition; Computer Science - Learning; Engineering::Computer science and engineering; Transfer Learning; Distance Metric Learning; Moe; DML
• Online Access: https://hdl.handle.net/10356/141947; https://doi.org/10.1109/TPAMI.2018.2824309

The goal of transfer learning is to improve the performance of target learning task by leveraging information (or transferring knowledge) from other related tasks. In this paper, we examine the problem of transfer distance metric learning (DML), which usually aims to mitigate the label information deficiency issue in the target DML. Most of the current Transfer DML (TDML) methods are not applicable to the scenario where data are drawn from heterogeneous domains. Some existing heterogeneous transfer learning (HTL) approaches can learn target distance metric by usually transforming the samples of source and target domain into a common subspace. However, these approaches lack flexibility in real-world applications, and the learned transformations are often restricted to be linear. This motivates us to develop a general flexible heterogeneous TDML (HTDML) framework. In particular, any (linear/nonlinear) DML algorithms can be employed to learn the source metric beforehand. Then the pre-learned source metric is represented as a set of knowledge fragments to help target metric learning. We show how generalization error in the target domain could be reduced using the proposed transfer strategy, and develop novel algorithm to learn either linear or nonlinear target metric. Extensive experiments on various applications demonstrate the effectiveness of the proposed method. Ministry of Education (MOE) Accepted version This research was supported in part by Singapore NRF2015ENC-GDCR01001-003, administrated via IMDA, NRF2015ENC-GBICRD001-012, administrated via BCA, DSAIR@NTU, and Australian Research Council Projects FL-170100117, DP-180103424, and DP-140102164.