Contextual road lane and symbol generation for autonomous driving

TL;DR

This paper introduces a generative adversarial network-based approach for lane detection and symbol generation in autonomous driving, improving robustness and performance over traditional methods especially in challenging conditions.

Contribution

It presents a novel generative model for lane and symbol detection that outperforms existing discriminative models and handles occlusions and faded scenarios effectively.

Findings

Outperforms state-of-the-art methods on BDD100K and Baidu ApolloScape datasets.

Robustly generates lanes in occluded and faded conditions.

Demonstrates improved accuracy and reliability in adverse scenarios.

Abstract

In this paper we present a novel approach for lane detection and segmentation using generative models. Traditionally discriminative models have been employed to classify pixels semantically on a road. We model the probability distribution of lanes and road symbols by training a generative adversarial network. Based on the learned probability distribution, context-aware lanes and road signs are generated for a given image which are further quantized for nearest class label. Proposed method has been tested on BDD100K and Baidu ApolloScape datasets and performs better than state of the art and exhibits robustness to adverse conditions by generating lanes in faded out and occluded scenarios.