Paper Reading AI Learner

Auto Seg-Loss: Searching Metric Surrogates for Semantic Segmentation

2020-10-15 17:59:08
Hao Li, Chenxin Tao, Xizhou Zhu, Xiaogang Wang, Gao Huang, Jifeng Dai

Abstract

tract: We propose a general framework for searching surrogate losses for mainstream semantic segmentation metrics. This is in contrast to existing loss functions manually designed for individual metrics. The searched surrogate losses can generalize well to other datasets and networks. Extensive experiments on PASCAL VOC and Cityscapes demonstrate the effectiveness of our approach. Code shall be released.

Abstract (translated)

URL

https://arxiv.org/abs/2010.07930

PDF

https://arxiv.org/pdf/2010.07930