Ripple-GAN: Lane Line Detection With Ripple Lane Line Detection Network and Wasserstein GAN

With artificial intelligence technology being advanced by leaps and bounds, intelligent driving has attracted a huge amount of attention recently in research development. In driving, lane line detection is fundamental but challenging task particularly under complex road conditions. this paper, we propose simple yet appealing network called Ripple Lane Line Detection Network (RiLLD-Net), to exploit quick connections gradient maps for effective learning features. RiLLD-Net can handle most common scenes detection. Then, order address scenarios such as occluded or lines, more powerful Ripple-GAN, integrating RiLLD-Net, confrontation training Wasserstein generative adversarial networks, multi-target semantic segmentation. Experiments show that, especially obscured Ripple-GAN produce superior performance other state-of-the-art methods.