YOLOR 5GBS

--- library_name: ultralytics pipeline_tag: object-detection tags: - yolo - yolov11 - object-detection - coco - 5g - beamforming - vibe - yolor --- # YOLOR-5GBS      <table> <tr> <td width="30%" valign="top"> <img src="5GBS.png" alt="YOLOR-5GBS — example 5G BS and LampPost detection" width="100%"> </td> <td valign="top"> YOLOv11x fine-tuned to detect outdoor RF infrastructure — `5G BS` (5G small cells) and `LampPost` with the 80 COCO classes. Part of the YOLOR detector family used for Stage 1 (camera priming) of the Look Once, Beam Twice mmWave V2X beam-management pipeline (SECON 2026). Data was captured in **Downtown [Lincoln, Nebraska](https://lincoln.ne.gov/), USA**. </td> </tr> </table> Reference implementation for the paper: > Avhishek Biswas\*, Apala Pramanik\*, Eylem Ekici, Mehmet C. Vuran. > *"Look Once, Beam Twice: Camera-Primed Real-Time Double-Directional mmWave Beam Management for Vehicular Connectivity."* (\*equal contribution) > > arXiv: <https://doi.org/10.48550/arXiv.2605.05071> <p align="center"> <img src="overview2_updated.png" alt="VIBE five-stage camera-primed beam-management pipeline" width="90%"> </p> ## Quick links - Paper (arXiv): <https://doi.org/10.48550/arXiv.2605.05071> - Code: <https://github.com/UNL-CPN-Lab/Look-Once-Beam-Twice> - Training pipeline: <https://github.com/UNL-CPN-Lab/Look-Once-Beam-Twice/tree/main/YOLOR_Training> | | | |---|---| | **Architecture** | YOLOv11x, 82-class output head (COCO 80 + 2 custom) | | **Initialization** | stock `yolo11x.pt` | | **Schedule** | 200 epochs, `cos_lr`, `close_mosaic=20`, `lr0=0.01` | | **Training data** | OutdoorDataset labeled subset — 4,107 train / 336 val / 362 test | | **Custom classes** | `5G BS` (id 80), `LampPost` (id 81) | | **Released checkpoint** | `last.pt` | ## Usage ```python from huggingface_hub import hf_hub_download from ultralytics import YOLO weights = hf_hub_download(repo_id="cpnlab/YOLOR-5GBS", filename="last.pt") model = YOLO(weights) results = model.predict("path/to/im...