Reproduce “RaSFNet: Ray-aligned and Scan-aware Fusion Network for LiDAR Semantic Segmentation” 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
Chuang Chen
RaSFNet segments LiDAR points using projected cells, point geometry around sensing rays, scan-order context, and range-guided exchange between points and cells. On SemanticKITTI validation, it reports 70.2% mIoU versus 67.6% for FRNet. Removing range-conditioned fusion lowers mIoU to 67.5% under identical training settings. The reported cross-method comparison is not a protocol-matched test.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.