RaDICaL: A Synchronized FMCW Radar, Depth, IMU and RGB Camera Data Dataset with Low-Level FMCW Radar Signals
Teck-Yian Lim*, Spencer Markowitz*, Minh N. Do
Within the autonomous driving community, radars are a significantly under-exploited sensor. This is despite radar's capability to operate in many conditions where traditional cameras fail, such as low lighting, bad weather, and occlusion. This work seeks to bridge this gap by providing the community with a diverse minimally processed FMCW radar dataset that is not only RGB-D aligned but also synchronized with IMU measurements in the presence of ego-motion. Moreover, having time-synchronized measurements allows for verification, automated labeling of the radar data, and opens the door for novel methods of fusing the data from a variety of sensors. Furthermore, our dataset provides diverse scenes with applications that extend beyond the automotive industry. All components to build our data collection system are off-the-shelf items and can be obtained within a $1000 budget. Consequently, we believe this dataset is relevant to fields spanning from automated systems, consumer electronics, and radar signal processing research. Finally, we demonstrated the ability to go beyond CFAR object detection with our dataset with a classification performance of 86% using the low-level radar signals captured by our dataset.
Article and dataset pending review. Stay tuned!
The dataset will be hosted here
We thank Texas Instruments for their support in this project.