AdaNFF: A new method for adaptive nonnegative multi-feature fusion to scene classification

作者：

Highlights：

• Adaptive nonnegative feature fusion framework is proposed for multi-feature fusion.

• Feature fusion boosting algorithm is proposed to improve classificaion performance.

• Dimension equilibirum strategy is proposed and utilized to optimize nonnegative feature factorization.

• The proposed methods achieve remarkable classification performance on scene image benchmarks.

摘要

•Adaptive nonnegative feature fusion framework is proposed for multi-feature fusion.•Feature fusion boosting algorithm is proposed to improve classificaion performance.•Dimension equilibirum strategy is proposed and utilized to optimize nonnegative feature factorization.•The proposed methods achieve remarkable classification performance on scene image benchmarks.

论文关键词：Scene classification,Adaptive feature fusion,Nonnegative matrix factorization,Feature fusion boosting

论文评审过程：Received 25 February 2021, Revised 15 September 2021, Accepted 24 October 2021, Available online 26 October 2021, Version of Record 2 November 2021.

论文官网地址：https://doi.org/10.1016/j.patcog.2021.108402