Reproduce “SelectOccFlow: Selective Spatiotemporal Aggregation for 3D Occupancy and Scene Flow Prediction” 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
Yuhang Wang, Kai Luo, Yuanfan Zheng, Kailun Yang
SelectOccFlow predicts 3D semantic occupancy and voxel motion by selectively aggregating image features, historical evidence, and directional voxel support. Image semantics guide sampling, voxel state guides history retrieval, and directional structure refines voxel features. On OpenOcc, its OccScore, combining RayIoU and scene-flow error, is 44.9 versus 43.1 for ALOcc; this measures benchmark prediction, not safer driving.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.