SPAN：面向单目 3D 目标检测的空间-投影对齐
SPAN: Spatial-Projection Alignment for Monocular 3D Object Detection
把解耦回归丢失的几何一致性先验显式补回来：用 3D 框点级空间对齐和 3D-2D 投影对齐作为可插拔的损失/约束，配合分层任务学习逐步引入，可直接挂到现有单目 3D 检测器上提升精度。
原文依据 3D-2D Projection Alignment ensures that the projected 3D box is aligned tightly within its corresponding 2D detection bounding box on the image plane
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论现有单目 3D 检测器通常采用解耦预测范式来驯服 3D 边界框显著的强非线性回归问题，即使用多个分支分别估计几何中心、深度、尺寸与旋转角。尽管这种解耦策略简化了学习过程，但它本质上忽略了不同属性之间的几何协同约束，导致缺乏几何一致性先验，从而性能次优。为解决该问题，我们提出了新颖的空间-投影对齐（SPAN），包含两个关键组件：(i) 空间点对齐在预测的 3D 边界框与真值 3D 边界框之间施加显式的全局空间约束，从而纠正由解耦属性回归引起的空间漂移；(ii) 3D-2D 投影对齐确保投影后的 3D 框在图像平面上紧密对齐于其对应的 2D 检测边界框之内，缓解了以往工作所忽视的投影错位问题。为保证训练稳定性，我们进一步引入分层任务学习策略，随着 3D 属性预测的逐步精化而渐进地纳入空间-投影对齐，防止早期阶段的误差在属性间传播。大量实验表明，所提方法可以轻松集成到任何已有的单目 3D 检测器中，并带来显著的性能提升。
英文原文摘要 1467 字符
Existing monocular 3D detectors typically tame the pronounced nonlinear regression of 3D bounding box through decoupled prediction paradigm, which employs multiple branches to estimate geometric center, depth, dimensions, and rotation angle separately.Although this decoupling strategy simplifies the learning process, it inherently ignores the geometric collaborative constraints between different attributes, resulting in the lack of geometric consistency prior, thereby leading to suboptimal performance. To address this issue, we propose novel Spatial-Projection Alignment (SPAN) with two pivotal components: (i). Spatial Point Alignment enforces an explicit global spatial constraint between the predicted and ground-truth 3D bounding boxes, thereby rectifying spatial drift caused by decoupled attribute regression. (ii). 3D-2D Projection Alignment ensures that the projected 3D box is aligned tightly within its corresponding 2D detection bounding box on the image plane, mitigating projection misalignment overlooked in previous works. To ensure training stability, we further introduce a Hierarchical Task Learning strategy that progressively incorporates spatial-projection alignment as 3D attribute predictions refine, preventing early stage error propagation across attributes. Extensive experiments demonstrate that the proposed method can be easily integrated into any established monocular 3D detector and delivers significant performance improvements.