SG-LDM: Semantic-Guided LiDAR Generation via Latent-Aligned Diffusion
This work addresses the need for diverse and controllable LiDAR data synthesis to enhance training for perception models in autonomous driving or robotics, representing a novel application rather than an incremental improvement.
The paper tackles the problem of generating high-fidelity LiDAR point clouds guided by semantic labels, achieving state-of-the-art performance and improving downstream segmentation tasks through a novel translation framework.
Lidar point cloud synthesis based on generative models offers a promising solution to augment deep learning pipelines, particularly when real-world data is scarce or lacks diversity. By enabling flexible object manipulation, this synthesis approach can significantly enrich training datasets and enhance discriminative models. However, existing methods focus on unconditional lidar point cloud generation, overlooking their potential for real-world applications. In this paper, we propose SG-LDM, a Semantic-Guided Lidar Diffusion Model that employs latent alignment to enable robust semantic-to-lidar synthesis. By directly operating in the native lidar space and leveraging explicit semantic conditioning, SG-LDM achieves state-of-the-art performance in generating high-fidelity lidar point clouds guided by semantic labels. Moreover, we propose the first diffusion-based lidar translation framework based on SG-LDM, which enables cross-domain translation as a domain adaptation strategy to enhance downstream perception performance. Systematic experiments demonstrate that SG-LDM significantly outperforms existing lidar diffusion models and the proposed lidar translation framework further improves data augmentation performance in the downstream lidar segmentation task.