LRR-Sim: Long-Range Radar Simulation Dataset

LRR-Sim is the first publicly available long-range radar dataset with annotated 3D point clouds extending up to 300 meters, filling a critical gap in radar research for highway and high-speed scenarios. It includes ~18K training and ~3K testing frames across 50 simulated highway scenes, with a total of ~160K annotated vehicles - including ~28K vehicles beyond 175m, where conventional sensors struggle. Scenes were created using the CARLA simulator for realistic world modeling and highway maps, with a high-fidelity radar simulation applied on top to generate 77GHz MIMO radar point clouds. All objects within ±55° azimuth and ±20° elevation are annotated with accurate 3D bounding boxes.

---

🧠 Key Features

• 📡 Realistic radar simulated dataset with range-Doppler-azimuth-elevation spectra.
• 🚗 Highway scenes with diverse vehicles (car, van, truck).
• 📏 Range up to 300m - the longest detection range among public radar datasets.
• 🎯 Precise 3D annotations for all vehicles within the radar field-of-view.
• ⚙️ Radar parameters based on high-end automotive sensors.
• 📦 Comes with a Python Dataset class for easy integration into ML pipelines.

---

🔗 Download

You can download the dataset using the links below:

• 📥 Download Radar Data (105 MB)
• 📥 Download Image Data (13.5 GB)

---

📁 Dataset Structure

Each scenario contains radar data collected at 20 FPS for 30 seconds, saved as individual frames:

lrrsim/
├── image_data/
│ ├── train/
│ │ ├── 00000.jpg
│ │ ├── 00001.jpg
│ │ └── ...
│ ├── test/
│ │ ├── 00000.jpg
│ │ ├── 00000.jpg
│ │ └── ...
├── radar_data/
│ ├── train/
│ │ ├── 00000.pkl
│ │ ├── 00001.pkl
│ │ └── ...
│ ├── test/
│ │ ├── 00000.pkl
│ │ ├── 00001.pkl
│ │ └── ...

Each radar file xxxxx.pkl file contains a dictionary:

• radar_pts (N × 8): Radar point cloud [x, y, z, doppler, intensity, frame_idx]
• ego_vel (3,): The ego velocity
• T_ego_local2global (4,4): convert from ego local coordinates to global coordinates
• timestamp : frame timestamp in sec
• targets_bboxes (K,7): GT bboxes [cx, cy, cz, ex, ey, ez, yaw(deg)]
• targets_vel (K,3): Target vehicles GT velocities
• target_ids (K,): Target vehicles unique ID
• target_is_visible (K,): Flag indicating if target vehicle is visible to sensor
• img: front camera image (optional)

---

📐 Radar Sensor Configuration

Parameter	Value
Max Range	300m
Range Resolution	0.15m
Azimuth FOV	±55°
Azimuth Resolution	1.2°
Elevation FOV	±20°
Elevation Resolution	2°
Doppler Range	[−80, +30] m/s
Doppler Resolution	0.13 m/s

---

🔧 Requirements

• torch
• numpy
• opencv-python
• matplotlib >= 3.6

---

🧰 Usage Example

from lrrsim import LRRSimDataset

DATA_PATH = "path/to/LRRSimDatase"
SPLIT = 'train' # 'test'

dset = LRRSimDataset(data_path=DATA_PATH, split=SPLIT, load_imgs=False)

idx = 0
data_dict = dset[idx]

show_radar_scene(data_dict)

---

📊 Dataset Statistics

🛣 50 highway driving scenarios (42 train / 8 test)

⏱ Approximately 30s each at 20 FPS (18172 frames for train, 3363 frames for test)

🚘 ~160K vehicle instances, ~28K beyond 175m

📦 Vehicle types include cars, trucks, and vans, with an average of 7.3 vehicles per frame

---

📜 Citation

If you use LRR-Sim in your research, please cite:

@inproceedings{doppdrive2025,
  title={DoppDrive: Doppler-Driven Temporal Aggregation for Improved Radar Object Detection},
  author={Yuval Haitman and Oded Bialer},
  booktitle={ICCV},
  year={2025}
}