ISAIR: Deep inpainted semantic aware image representation for background subtraction

作者：

Highlights：

• New image representation based on detecting and in painting inherently moving objects.

• Benefiting from advantages of traditional and deep learning approaches simultaneously.

• Robust background modeling in different environments without tuning parameters.

• End-to-end framework outperforms state-of-the-art background subtraction techniques.

摘要

•New image representation based on detecting and in painting inherently moving objects.•Benefiting from advantages of traditional and deep learning approaches simultaneously.•Robust background modeling in different environments without tuning parameters.•End-to-end framework outperforms state-of-the-art background subtraction techniques.

论文关键词：Image representation,Background subtraction,Deep learning

论文评审过程：Received 11 October 2020, Revised 26 May 2022, Accepted 20 June 2022, Available online 30 June 2022, Version of Record 4 July 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.117947