EL-GAN: Embedding Loss Driven Generative Adversarial Networks for Lane Detection

TL;DR

EL-GAN introduces an embedding loss in a GAN framework to improve lane detection by producing more realistic segmentation outputs, reducing reliance on complex post-processing, and achieving high accuracy on the TuSimple benchmark.

Contribution

The paper proposes EL-GAN, a novel GAN-based approach using embedding loss for more stable training and improved lane detection performance.

Findings

Achieves over 96% accuracy on TuSimple benchmark

Produces more realistic and structure-preserving lane segmentation

Simplifies post-processing compared to traditional methods

Abstract

Convolutional neural networks have been successfully applied to semantic segmentation problems. However, there are many problems that are inherently not pixel-wise classification problems but are nevertheless frequently formulated as semantic segmentation. This ill-posed formulation consequently necessitates hand-crafted scenario-specific and computationally expensive post-processing methods to convert the per pixel probability maps to final desired outputs. Generative adversarial networks (GANs) can be used to make the semantic segmentation network output to be more realistic or better structure-preserving, decreasing the dependency on potentially complex post-processing. In this work, we propose EL-GAN: a GAN framework to mitigate the discussed problem using an embedding loss. With EL-GAN, we discriminate based on learned embeddings of both the labels and the prediction at the same time. This results in more stable training due to having better discriminative information, benefiting from seeing both `fake' and `real' predictions at the same time. This substantially stabilizes the adversarial training process. We use the TuSimple lane marking challenge to demonstrate that with our proposed framework it is viable to overcome the inherent anomalies of posing it as a semantic segmentation problem. Not only is the output considerably more similar to the labels when compared to conventional methods, the subsequent post-processing is also simpler and crosses the competitive 96% accuracy threshold.