Attribute distance for fast classification systems
作者：
Highlights：
• An attribute distance is proposed for estimating the information of individual images.
• The attribute distance is estimated on some discriminative attributes instead of all.
• The attribute distance can be used to speed up the training process.
• The time expense for constructing the active classification system is reduced.
• The proposed method can obtain a more accurate classifier than other baselines.
摘要
•An attribute distance is proposed for estimating the information of individual images.•The attribute distance is estimated on some discriminative attributes instead of all.•The attribute distance can be used to speed up the training process.•The time expense for constructing the active classification system is reduced.•The proposed method can obtain a more accurate classifier than other baselines.
论文关键词：Classification systems,Active feedback,Attribute selection,Similarity evaluation
论文评审过程：Received 26 September 2017, Revised 13 February 2018, Accepted 21 March 2018, Available online 29 March 2018, Version of Record 24 April 2018.
论文官网地址：https://doi.org/10.1016/j.eswa.2018.03.049
