Dr.Occ：面向自动驾驶的深度与区域引导环视相机三维占用预测
Dr.Occ: Depth- and Region-Guided 3D Occupancy from Surround-View Cameras for Autonomous Driving
用 MoGe-2 的稠密深度先验监督 2D-to-3D 视角变换，可缓解纯视觉 BEV/占用方法中体素特征与几何错位的问题；区域引导的 MoE 专家分配则为处理空间类别不平衡提供了可复用的思路。
原文依据 a depth-guided 2D-to-3D View Transformer (D-VFormer) that effectively leverages high-quality dense depth cues from MoGe-2
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论三维语义占用预测对自动驾驶感知至关重要，能够提供全面的几何场景理解与语义识别。然而，现有方法由于缺乏像素级精确的深度估计，在视角变换中存在几何错位问题；同时语义类别呈现强烈的空间各向异性，导致严重的空间类别不平衡。为解决上述挑战，我们提出 Dr.Occ，一个深度与区域引导的占用预测框架。具体而言，我们引入深度引导的二维到三维视角 Transformer（D-VFormer），有效利用来自 MoGe-2 的高质量稠密深度线索构建可靠的几何先验，从而实现体素特征的精确几何对齐。此外，受混合专家（MoE）框架启发，我们提出区域引导的专家 Transformer（R/R-EFormer），自适应地分配区域专属专家以关注不同空间区域，有效应对空间语义差异。由此，两个组件形成互补贡献：深度引导确保几何对齐，区域专家增强语义学习。在 Occ3D--nuScenes 基准上的实验表明，在纯视觉完整设置下，Dr.Occ 将强基线 BEVDet4D 提升 7.43% 平均交并比和 3.09% IoU。
英文原文摘要 1338 字符
3D semantic occupancy prediction is crucial for autonomous driving perception, offering comprehensive geometric scene understanding and semantic recognition. However, existing methods struggle with geometric misalignment in view transformation due to lack of pixel-level accurate depth estimation, and severe spatial class imbalance where semantic categories exhibit strong spatial anisotropy. To address these challenges, we propose Dr.Occ, a depth- and region-guided occupancy prediction framework. Specifically, we introduce a depth-guided 2D-to-3D View Transformer (D-VFormer) that effectively leverages high-quality dense depth cues from MoGe-2 to construct reliable geometric priors, thereby enabling precise geometric alignment of voxel features. Moreover, inspired by the Mixture-of-Experts (MoE) framework, we propose a region-guided Expert Transformer (R/R-EFormer) that adaptively allocates region-specific experts to focus on different spatial regions, effectively addressing spatial semantic variations. Thus, the two components make complementary contributions: depth guidance ensures geometric alignment, while region experts enhance semantic learning. Experiments on the Occ3D--nuScenes benchmark demonstrate that Dr.Occ improves the strong baseline BEVDet4D by 7.43% mIoU and 3.09% IoU under the full vision-only setting.