用少量标注实现高精度3D语义占据预测，兼顾效率与性能。
OccLE: Label-Efficient 3D Semantic Occupancy Prediction
- 分步学习语义与几何特征，再通过双分支融合提升精度。
- 仅需10%标注即可在SemanticKITTI等数据集上达到领先效果。
- 适合自动驾驶场景的低成本高精度环境理解任务。
3D语义占据预测为自动驾驶感知提供了直观高效的场景理解方式，但现有方法或依赖全监督（需昂贵体素级标注），或依赖自监督（引导有限，性能不足）。为此，我们提出OccLE，一种标签高效3D语义占据预测方法，以图像和激光雷达为输入，在仅使用少量体素标注的情况下保持高性能。核心思路是将语义与几何学习解耦：语义分支利用2D基础模型生成对齐的伪标签，用于2D与3D语义学习；几何分支基于图像与激光雷达的内在互补性，通过跨平面协同实现半监督几何学习。最终通过双Mamba融合语义-几何特征网格，并采用散射累积投影机制，以对齐伪标签监督未标注区域。实验表明，OccLE在SemanticKITTI和Occ3D-nuScenes数据集上仅需10%体素标注即达到竞争力表现。
原文摘要 · Abstract (English)
3D semantic occupancy prediction offers an intuitive and efficient scene understanding and has attracted significant interest in autonomous driving perception. Existing approaches either rely on full supervision, which demands costly voxel-level annotations, or on self-supervision, which provides limited guidance and yields suboptimal performance. To address these challenges, we propose OccLE, a Label-Efficient 3D Semantic Occupancy Prediction that takes images and LiDAR as inputs and maintains high performance with limited voxel annotations. Our intuition is to decouple the semantic and geometric learning tasks and then fuse the learned feature grids from both tasks for the final semantic occupancy prediction. Therefore, the semantic branch distills 2D foundation model to provide aligned pseudo labels for 2D and 3D semantic learning. The geometric branch integrates image and LiDAR inputs in cross-plane synergy based on their inherency, employing semi-supervision to enhance geometry learning. We fuse semantic-geometric feature grids through Dual Mamba and incorporate a scatter-accumulated projection to supervise unannotated prediction with aligned pseudo labels. Experiments show that OccLE achieves competitive performance with only 10\% of voxel annotations on the SemanticKITTI and Occ3D-nuScenes datasets. The code will be publicly released on https://github.com/NerdFNY/OccLE
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。