QueryOcc：面向 3D 语义占用的查询式自监督
QueryOcc: Query-based Self-Supervision for 3D Semantic Occupancy
用跨相邻帧采样的 4D 时空查询直接对连续 3D 占用做自监督，配合收缩式场景表示在恒定显存下覆盖远距离区域，可替代基于 2D 渲染一致性或离散体素的监督方式，用于相机/激光雷达输入的占用学习。
原文依据 a query-based self-supervised framework that learns continuous 3D semantic occupancy directly through independent 4D spatio-temporal queries
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论从图像中学习三维场景的几何与语义是计算机视觉的核心挑战，也是自动驾驶的关键能力。由于大规模三维标注代价高昂，近期工作探索直接从传感器数据进行自监督学习而无需人工标注。现有方法要么依赖二维渲染一致性，三维结构只能隐式涌现；要么依赖由累积激光雷达点云离散化得到的体素网格，限制了空间精度与可扩展性。本文提出 QueryOcc，一个基于查询的自监督框架，通过在相邻帧间采样的独立四维时空查询直接学习连续的 3D 语义占用。该框架既支持来自视觉基础模型的伪点云监督，也支持原始激光雷达数据监督。为在恒定显存下实现远距离监督与推理，我们引入一种收缩式场景表示，在保留近场细节的同时平滑压缩远处区域。在自监督 Occ3D-nuScenes 基准上，QueryOcc 的语义 RayIoU 较此前基于相机的方法提升 26%，运行速度达 11.6 FPS，表明直接的四维查询监督能够实现强自监督占用学习。
英文原文摘要 1269 字符
Learning 3D scene geometry and semantics from images is a core challenge in computer vision and a key capability for autonomous driving. Since large-scale 3D annotation is prohibitively expensive, recent work explores self-supervised learning directly from sensor data without manual labels. Existing approaches either rely on 2D rendering consistency, where 3D structure emerges only implicitly, or on discretized voxel grids from accumulated lidar point clouds, limiting spatial precision and scalability. We introduce QueryOcc, a query-based self-supervised framework that learns continuous 3D semantic occupancy directly through independent 4D spatio-temporal queries sampled across adjacent frames. The framework supports supervision from either pseudo-point clouds derived from vision foundation models or raw lidar data. To enable long-range supervision and reasoning under constant memory, we introduce a contractive scene representation that preserves near-field detail while smoothly compressing distant regions.QueryOcc surpasses previous camera-based methods by 26% in semantic RayIoU on the self-supervised Occ3D-nuScenes benchmark while running at 11.6 FPS, demonstrating that direct 4D query supervision enables strong self-supervised occupancy learning.