SDNet: LiDAR Semantic Scene Completion with Sparse-Dense Fusion and Input-Aware Label Refinement
Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (4) · 2381-2389
DOI: 10.1609/aaai.v40i4.37223 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Tingming Bai und 5 weiteren Autor:innen.
Abstract (Originalsprache)
LiDAR Semantic Scene Completion (SSC) in autonomous driving requires predicting both dense occupancy and semantic labels from sparse input point cloud. Existing methods typically adopt cascaded architecture for feature dilation and semantic abstraction, which blurs distinctive geometric patterns and reduces feature discriminability. Moreover, given an input, conventional processing of the ground truth labels overlooks voxel predictability in the target, resulting in ill-posed supervision and discards informative voxels. To address these limitations, we propose Sparse-Dense Net (SDNet), a dual-branch architecture that processes the input points through parallel sparse and dense encoders. The complementary features are aligned and fused using a Sparse Dense Feature Fusion (SDFF) module and further refined by a Feature Propagation (FP) module. Additionally, we introduce an input-aware label refinement strategy, including Sparse-Guided Filtering (SGF) to filter unpredictable targets and Ignored Voxel Recycling (IVR) to leverage informative ignored voxels for auxiliary supervision. These innovations enhance both feature learning and label quality. Extensive experiments on SemanticKITTI and nuScenes OpenOccupancy datasets validate the effectiveness of our approach, with SDNet achieving state-of-the-art performance on both datasets and ranking 1st on the official SemanticKITTI benchmark with 42.1 mIoU, outperforming the previous best by 4.2 (+11.1\%).
Zitieren
APA
Bai, T., Xiang, Z., Xu, P., Pu, T., Wang, K., & Liu, E. (2026). SDNet: LiDAR Semantic Scene Completion with Sparse-Dense Fusion and Input-Aware Label Refinement. Proceedings of the AAAI Conference on Artificial Intelligence, 40(4), 2381-2389. https://doi.org/10.1609/aaai.v40i4.37223
Vancouver
Bai T, Xiang Z, Xu P, Pu T, Wang K, Liu E. SDNet: LiDAR Semantic Scene Completion with Sparse-Dense Fusion and Input-Aware Label Refinement. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(4):2381-2389. doi:10.1609/aaai.v40i4.37223
BibTeX
@article{Bai2026,
author = {Bai, Tingming and Xiang, Zhiyu and Xu, Peng and Pu, Tianyu and Wang, Kai and Liu, Eryun},
title = {SDNet: LiDAR Semantic Scene Completion with Sparse-Dense Fusion and Input-Aware Label Refinement},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2026},
volume = {40},
number = {4},
pages = {2381--2389},
doi = {10.1609/aaai.v40i4.37223}
}
EUVIMED