Detection of panoramic vision pedestrian based on deep learning

作者：

Highlights：

• Anchor-free and one-stage detectors were used in pedestrian detection technology.

• The CSP of pedestrian detector makes the training process more robust.

• A novel method was proposed to predict width–compressing width.

• An anchor-free and one-stage detector can still have high accuracy.

摘要

•Anchor-free and one-stage detectors were used in pedestrian detection technology.•The CSP of pedestrian detector makes the training process more robust.•A novel method was proposed to predict width–compressing width.•An anchor-free and one-stage detector can still have high accuracy.

论文关键词：Pedestrian detection,Panoramic vision,Switchable normalization,Convolutional neural networks,Deep learning

论文评审过程：Received 30 May 2020, Revised 12 July 2020, Accepted 21 July 2020, Available online 3 August 2020, Version of Record 20 August 2020.

论文官网地址：https://doi.org/10.1016/j.imavis.2020.103986