YOLOR

YOLOR is a fine-tuned object detection model for BS identification for beam initialization to detect all five YOLOR custom classes — radio, 5G BS, LampPost, mmWave radio, streetlight — in one inference pass. The combined release model of the YOLOR detector family used for the Look Once, Beam Twice mmWave V2X beam-management pipeline (SECON 2026).

Source hardware and models

Model	Source hardware / location	Hugging Face
YOLOR-radio	Sivers Semiconductors 60 GHz mmWave Radio frontends (EVK06002)	cpnlab/YOLOR-radio
YOLOR-5GBS	5G small cells + co-located lamp/utility poles, captured in Downtown Lincoln, Nebraska, USA	cpnlab/YOLOR-5GBS
YOLOR-comm-mmWave	Terragraph Sounders from Meta, deployed in indoor commercial spaces	cpnlab/YOLOR-comm-mmWave
YOLOR-Streetlights	Urban streetlights on the University of Nebraska–Lincoln campus	cpnlab/YOLOR-Streetlights
YOLOR (unified)	Union of all four sources above	this card

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

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, 85-class output head (COCO 80 + 5 custom)
Initialization	stock yolo11x.pt
Schedule	200 epochs, cos_lr, close_mosaic=20, lr0=0.01
Training data	union of all four YOLOR source domains (cots + outdoor + commercial + streetlight), ~10,800 custom train frames + 8,000 COCO replay
Custom classes	radio (80), 5G BS (81), LampPost (82), mmWave radio (83), streetlight (84)
Released checkpoint	last.pt

Usage

from huggingface_hub import hf_hub_download
from ultralytics import YOLO

weights = hf_hub_download(repo_id="cpnlab/YOLOR", filename="last.pt")
model = YOLO(weights)
results = model.predict("path/to/image.jpg", conf=0.25)

Class indices in returned detections:

• 0–79 — the 80 standard COCO classes
• 80 — radio
• 81 — 5G BS
• 82 — LampPost
• 83 — mmWave radio
• 84 — streetlight

Training data and Code

Code and Data: https://github.com/UNL-CPN-Lab/Look-Once-Beam-Twice

Citation

@inproceedings{biswas2026look,
  title     = {Look Once, Beam Twice: Camera-Primed Real-Time Double-Directional
               mmWave Beam Management for Vehicular Connectivity},
  author    = {Biswas, Avhishek and Pramanik, Apala and Ekici, Eylem and Vuran, Mehmet C.},
  booktitle = {Proc. IEEE SECON},
  year      = {2026}
}

Paper: https://doi.org/10.48550/arXiv.2605.05071

Contact

For questions about this model or the paper, contact the corresponding authors:

• Avhishek Biswas — abiswas3@huskers.unl.edu
• Apala Pramanik — apramanik2@huskers.unl.edu

Acknowledgments

Developed at the Cyber Physical Networking (CPN) Lab, School of Computing, University of Nebraska–Lincoln, in collaboration with The Ohio State University. Thanks to Sivers Semiconductors, Ettus Research, and the open-source Ultralytics, PyTorch, and Ettus UHD communities.