185fdf627eaae2abab36205dcd19b817-Supplemental-Datasets_and_Benchmarks.pdf

Appendix The appendix is organized as follows. We also provide details of the annotation/calibration process and the baseline neural networks (NNs) in Section D and E, respectively. We discuss results regarding each weather condition and consideration of the K-Radar dataset as a pre-training dataset for other Radar tensor datasets in Section F and G, respectively. Finally, we introduce details of devkits and list relevant URLs to help with understanding the content of the paper in Section H and I, respectively. A.1 Additional samples of the K-Radar dataset and explanation of LPCs for each weather condition In the sleet (Figure 8-(e)) or heavy snow (Figure 8-(g)) condition, the Lidar point cloud (LPC) measurements of some objects ahead are lost when the ego-vehicle is driving.