To investigate the effectiveness of a multi-task neural network, we deployed a YOLOP[19]model capable of performing multiple tasks simultaneously, including object detection, drivable area segmentation, andlane
segmentation extends semantic segmentation by not only classifying pixels but also distinguishing between different instances within the same category, thus providing more accurate recognition results. Panoptic segmentation combines the strengths of both semantic and instance segmentation, thus offering a ...