ELM based smile detection using Distance Vector

作者：

Highlights：

• We verified the effectiveness of mouth deformation as an indicator of a smile.

• We proposed a novel feature descriptor for smile detection.

• The proposed method outperforms other methods on accuracy and effectiveness.

• We improved GENKI-4K benchmark by relabeling and removing duplications.

摘要

•We verified the effectiveness of mouth deformation as an indicator of a smile.•We proposed a novel feature descriptor for smile detection.•The proposed method outperforms other methods on accuracy and effectiveness.•We improved GENKI-4K benchmark by relabeling and removing duplications.

论文关键词：Smile detection,Pair-wise Distance Vector,Extreme Learning Machine

论文评审过程：Received 28 February 2017, Revised 7 February 2018, Accepted 18 February 2018, Available online 20 February 2018, Version of Record 27 February 2018.

论文官网地址：https://doi.org/10.1016/j.patcog.2018.02.019