Railway Obstacle Intrusion Detection Based on Convolution Neural Network Multitask Learning

The detection of train obstacle intrusion is very important for the safe running trains. In this paper, we design a multitask model to warn detected target obstacles in railway scenes. addition, multiobjective optimization algorithm that performs with different task complexity. Through shared structure reparameterized backbone network, our learning utilizes resources effectively. Our work achieves competitive results on both object and line detection, excellent inference time performance (50 FPS). first introduce approach realize assisted-driving function scene.