A ship pilot on-duty state recognition model based SlowFast network and attention mechanism

The unsuitable state of ship pilots on duty is a potentially unsafe factor in ship navigation. It has become an important issue of water traffic safety to identify the unsuitable state of pilots, reduce water traffic accidents caused by human factors and improve the safety supervision ability of ship navigation. Aiming at the problems of complex lighting conditions, non-fixed driver trajectory and easy occlusion in ship cockpit scenes, a behavior recognition model based on SlowFast network and EPSA attention mechanism is proposed.