Xinying Hong · Siyu Li · Kang Zeng · Hao Shi · Bomin Peng · Kailun Yang · Zhiyong Li

Abstract: Bird's Eye View (BEV) perception technology is crucial for autonomous driving, as it generates top-down 2D maps for environment perception, navigation, and decision-making. Nevertheless, the majority of current BEV map generation studies focusing on visual map generation lack depth-aware reasoning capabilities. They exhibit limited efficacy in managing occlusions and handling complex environments, with a notable decline in perceptual performance under adverse weather conditions or low-light scenarios. Therefore, this paper proposes TS-CGNet, which leverages Temporal-Spatial fusion with Centerline-Guided diffusion. This visual framework, grounded in prior knowledge, is designed for integration into any existing network for building BEV maps. Specifically, this framework is decoupled into three parts: Local mapping system involves the initial generation of semantic maps using purely visual information; The Temporal-Spatial Aligner Module (TSAM) integrates historical information into mapping generation by applying transformation matrices; The Centerline-Guided Diffusion Model (CGDM) is a prediction module based on the diffusion model. CGDM incorporates centerline information through spatial-attention mechanisms to enhance semantic segmentation reconstruction. We construct BEV semantic segmentation maps by our methods on the public nuScenes and the robustness benchmarks under various corruptions. Our method improves 1.90%, 1.73%, and 2.87% for perceived ranges of 60×30m, 120×60m, and 240×60m in the task of BEV HD mapping. TS-CGNet attains an improvement of 1.92% for perceived ranges of 100×100m in the task of BEV semantic mapping. Moreover, TS-CGNet achieves an average improvement of 2.92% in detection accuracy under varying weather conditions and sensor interferences in the perception range of 240×60m.

Here is a demonstration of the qualitative results in various perceived ranges:

• Release the code.
• Release the arxiv preprint.

If our work is helpful to you, please consider citing us by using the following BibTeX entry:

@misc{hong2025tscgnettemporalspatialfusionmeets,
      title={TS-CGNet: Temporal-Spatial Fusion Meets Centerline-Guided Diffusion for BEV Mapping}, 
      author={Xinying Hong and Siyu Li and Kang Zeng and Hao Shi and Bomin Peng and Kailun Yang and Zhiyong Li},
      year={2025},
      eprint={2503.02578},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2503.02578}, 
}