EFNet: Enhancement-Fusion Network for Semantic Segmentation

作者：

Highlights：

• We present the EFNet for semantic segmentation, leading to significant improvement.

• It works by enhancing an image for more features to facilitate the segmentation.

• It can be used highly flexibly as an “add-on” of an existing segmentation network.

摘要

•We present the EFNet for semantic segmentation, leading to significant improvement.•It works by enhancing an image for more features to facilitate the segmentation.•It can be used highly flexibly as an “add-on” of an existing segmentation network.

论文关键词：image enhancement,feature fusion,semantic segmentation,CNN

论文评审过程：Received 3 August 2020, Revised 3 April 2021, Accepted 7 May 2021, Available online 20 May 2021, Version of Record 28 May 2021.

论文官网地址：https://doi.org/10.1016/j.patcog.2021.108023