Learning Orientation Field for OSM-Guided Autonomous Navigation
Paper • 2503.18276 • Published
Learning Orientation Field for OSM-Guided Autonomous Navigation
This repository contains the paper's model checkpoint and accompanying files.
- Paper: https://arxiv.org/abs/2503.18276
- Code and usage: https://github.com/IMRL/OriField
- Training data: https://huggingface.co/datasets/yuminhhuang/OriField-raw-derived-bev-kitti
- Model: https://huggingface.co/yuminhhuang/OriField
The output.example/ directory contains 3 files (53,723,937 bytes). Files keep their original relative paths.
Model files: output.example/config.yaml, output.example/model_final.pth, and output.example/inference/model.onnx.
Download
hf download yuminhhuang/OriField --local-dir .
See the code repository for training and inference instructions. This release does not assign a new license to upstream data; consult the original KITTI/SemanticKITTI and other upstream terms where applicable.