Adaptive appearance modeling via hierarchical entropy analysis over multi-type features
作者：
Highlights：
• A framework to address the challenge of multi-feature based appearance modeling.
• Defining feature-independent information entropy as a unified criterion.
• A multi-feature selection random forest model.
• A sparse codebook model based on the “maximum discriminability”.
• A hierarchical maximum entropy model.
摘要
•A framework to address the challenge of multi-feature based appearance modeling.•Defining feature-independent information entropy as a unified criterion.•A multi-feature selection random forest model.•A sparse codebook model based on the “maximum discriminability”.•A hierarchical maximum entropy model.
论文关键词：Description model,Adaptive feature selection,Random forest,Hierarchical maximum entropy model,Image classification
论文评审过程：Received 11 February 2018, Revised 7 July 2019, Accepted 13 September 2019, Available online 19 September 2019, Version of Record 25 September 2019.
论文官网地址：https://doi.org/10.1016/j.patcog.2019.107059
