Dual-Radar Synthetic Dataset for Static Short-Range Object Detection

This dataset provides synthetic radar data generated using a raytracing-based simulation of a static short-range scenario. The simulated scene includes multiple objects in a semi-structured environment, such as a human and a forklift. Two radar sensors (Radar A and Radar B) observe the scene from different positions and angles, allowing for multi-perspective evaluation and sensor fusion analysis.

The dataset includes:

• Radar range-Doppler data for Radar A and Radar B (stored in configA.h5 and configB.h5)
• Python scripts for dataset generation and processing (generate_radar_dataset.py, radar_utils.py)
• Ground truth object information in yolo format for detection tasks
• Configuration parameters for radar signal processing, antenna setup, interference modeling, and sensor placement

The radar parameters in configB differ between Radar A and B in FFT size, resolution, and tilt angles, providing realistic variation for multi-sensor perception research. The dataset is suitable for training and evaluating radar-based object detection models (e.g., YOLO) and dual-radar fusion techniques.