Online Active Proposal Set Generation for weakly supervised object detection

作者：

Highlights：

• Weakly supervised detection methods are affected by redundant negative proposals.

• We propose an online proposal set generation algorithm to removes negative proposals.

• Our method improves the detection accuracy without adding much computation.

• Our method combines the two-step training process into one-step.

摘要

•Weakly supervised detection methods are affected by redundant negative proposals.•We propose an online proposal set generation algorithm to removes negative proposals.•Our method improves the detection accuracy without adding much computation.•Our method combines the two-step training process into one-step.

论文关键词：Weakly supervised learning,Object detection,Proposal sampling

论文评审过程：Received 3 December 2020, Revised 23 August 2021, Accepted 9 November 2021, Available online 23 November 2021, Version of Record 10 January 2022.

论文官网地址：https://doi.org/10.1016/j.knosys.2021.107726