TLSD: Breaking the Limit of Topological Lane Mapping with Graph Knowledge and Distance Awareness

Abstract

High-Definition (HD) maps are essential for both Advanced Driver-Assistance Systems (ADAS) and autonomous driving. However, offline HD map construction remains costly and challenging to maintain due to the dynamic nature of real-world environments. Consequently, online HD map generation using onboard sensors has become a key area of research. Despite recent advancements, existing deep learning-based methods often provide inaccurate output even using computationally heavy architectures, limiting their practicality for real-world applications. We introduce TLSD, an efficient end-to-end neural network that generates HD maps, incorporating both topological and geometric road information. To enhance both accuracy and efficiency, we introduce four key innovations: (1) an iterative refinement scheme within the decoder to progressively improve map predictions, (2) a group-wise one-to-many assignment strategy that accelerates training convergence, (3) a graph neural network (GNN) module that integrates lane segment coordinates for improved spatial reasoning, and (4) a distance-aware topological post-processing method that enhances the quality of connectivity outputs. We performed extensive experiments % on the widely used OpenLane-V2 benchmark and showed that TLSD achieves a significant improvement in OLUS score compared to existing methods, setting a new state-of-the-art benchmark, producing accurate HDMaps, and a connectivity graph. In particular, TLSD outperforms previous methods on the lane segment perception task (+3.13 in OLUS) and the lane centerline perception task (+3.20 in OLS), demonstrating superior performance in lane-based HD map generation. In addition, we introduce an efficient version, eTLSD, which incorporates a lightweight ResNet-18 backbone and still achieves competitive results, outperforming previous ResNet-50-based methods.