Hand gesture understanding by weakly-supervised fusing shallow/deep image attributes

作者：

Highlights：

• A novel fully-automatic and quality-guided ranking algorithm is proposed.

• A weakly-supervised learning paradigm encodes multiple image-level attributes.

• A comprehensive hand gesture recognition experiment is conducted.

摘要

•A novel fully-automatic and quality-guided ranking algorithm is proposed.•A weakly-supervised learning paradigm encodes multiple image-level attributes.•A comprehensive hand gesture recognition experiment is conducted.

论文关键词：Weakly-supervised,Finger-level metric,Shallow/deep,Feature fusion,Hand gesture

论文评审过程：Received 3 July 2019, Revised 9 December 2019, Accepted 21 December 2019, Available online 24 December 2019, Version of Record 9 January 2020.

论文官网地址：https://doi.org/10.1016/j.image.2019.115760