Reproduce “Channel-Wise Attention-Based Network for Self-Supervised Monocular Depth Estimation” 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
Jiaxing Yan, Hong Zhao, Penghui Bu, YuSheng Jin
CADepth-Net adapts self-supervised monocular depth prediction by adding channel-wise context aggregation in its encoder and feature recalibration during decoding. It trains from geometry using photometric agreement, without dense depth labels. On KITTI’s Eigen split, it reports Abs Rel of 0.105 and δ<1.25 of 0.892, versus Monodepth2’s 0.115 and 0.877, with median ground-truth scaling and an 80-meter cap.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.