3.1.1. Training Experimental Platform

Considering the limited computing resources and computing power of NVIDIA Jetson TX2, we used the workstation with the GPU model of NVIDIA RTX3090, CPU model of i7-10700, and memory size of 32 G to carry out the training part of the experiment. PyTorch 1.7.0 [50] based on the Python 3.8 language was adopted as the framework of our algorithm. We also used CUDA11.1 in our experiments to call the GPU for training acceleration. Subsequently, we transplanted the trained model into the NVIDIA Jetson TX2.
