Reproduce “Coarse-to-fine crack cue for robust crack detection” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Zelong Liu, Yuliang Gu, Zhichao Sun, Huachao Zhu, Xin Xiao, Bo Du, Laurent Najman, Yongchao Xu
The authors introduce CrackCue, a cue-guided segmentation method for pavement cracks. It estimates a crack-free background using pooling and reconstruction, then concatenates the image-background difference with the input for segmentation. In U-Net cross-dataset tests, mean ODS rose from 0.870 to 0.905 across three datasets. This result is specific to that protocol; very wide cracks may produce a weaker cue.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.