SED: Searching Enhanced Decoder with switchable skip connection for semantic segmentation

Xian Zhang, Zhibin Quan, Qiang Li, Dejun Zhu, Wankou Yang

Abstract: Highlights•For the task of image semantic segmentation, we propose a gradient-based, pre-trainable neural network architecture search framework SED. In this paper we simultaneously considering decoder and skip connection search. Our method maximizes the advantages of NAS and pre- trained backbone.•SED can compress the retraining iterations to several thousands. The whole searching, pruning, retraining process can be compressed into 1 day. Furthermore, after searching on Cityscapes, the searched network architecture can achieve 80.2% mIoU.