TY - JOUR

T1 - Zero-shot multi-label learning via label factorisation

AU - Shao, H.

AU - Guo, Y.

AU - Ding, G.

AU - Han, J.

PY - 2019/3/1

Y1 - 2019/3/1

N2 - This study considers the zero-shot learning problem under the multi-label setting where each test sample is associated with multiple labels that are unseen in training data. The authors propose a novel learning framework based on label factorisation for this problem. Specifically, the authors' framework takes three key issues into consideration and addresses them in a unified way. The first is knowledge transfer that utilises information from seen classes to build recognition models for unseen classes. The second is label correlation which means that labels which have different semantics may co-occur frequently. This is an important issue in multi-label learning. The authors propose to learn a shared latent space by label factorisation and use the label semantics as the decoding function, which can address both issues. The third is the predictability which requires the learned latent space to be strongly related to the visual features. It is guaranteed by incorporating a regression model into the learning framework. The authors derive two specific formulations from the general framework and propose the corresponding learning algorithms. The authors conducted extensive experiments on three multi-label data sets. The results demonstrated the effectiveness.

AB - This study considers the zero-shot learning problem under the multi-label setting where each test sample is associated with multiple labels that are unseen in training data. The authors propose a novel learning framework based on label factorisation for this problem. Specifically, the authors' framework takes three key issues into consideration and addresses them in a unified way. The first is knowledge transfer that utilises information from seen classes to build recognition models for unseen classes. The second is label correlation which means that labels which have different semantics may co-occur frequently. This is an important issue in multi-label learning. The authors propose to learn a shared latent space by label factorisation and use the label semantics as the decoding function, which can address both issues. The third is the predictability which requires the learned latent space to be strongly related to the visual features. It is guaranteed by incorporating a regression model into the learning framework. The authors derive two specific formulations from the general framework and propose the corresponding learning algorithms. The authors conducted extensive experiments on three multi-label data sets. The results demonstrated the effectiveness.

KW - Factorization

KW - Knowledge management

KW - Learning systems

KW - Regression analysis

KW - Semantics

KW - Knowledge transfer

KW - Label correlations

KW - Learning frameworks

KW - Learning problem

KW - Multi-label learning

KW - Multiple labels

KW - Recognition models

KW - Regression model

KW - Learning algorithms

U2 - 10.1049/iet-cvi.2018.5131

DO - 10.1049/iet-cvi.2018.5131

M3 - Journal article

VL - 13

SP - 117

EP - 124

JO - IET Computer Vision

JF - IET Computer Vision

SN - 1751-9632

IS - 2

ER -