Toward achieving robust low-level and high-level scene parsing

• Main Authors: Shuai, Bing, Ding, Henghui, Liu, Ting, Wang, Gang, Jiang, Xudong
• Other Authors: School of Electrical and Electronic Engineering
• Format: Article
• Language: English
• Published: 2020
• Subjects: Engineering::Electrical and electronic engineering; Scene Parsing; Convolution Neural Network
• Online Access: https://hdl.handle.net/10356/142866
• Institution: Nanyang Technological University

In this paper, we address the challenging task of scene segmentation. We first discuss and compare two widely used approaches to retain detailed spatial information from pretrained CNN - "dilation" and "skip". Then, we demonstrate that the parsing performance of "skip"...