UniBEVFusion：用于3D目标检测的统一雷达-视觉BEVFusion
UniBEVFusion: Unified Radar-Vision BEVFusion for 3D Object Detection
AI总结：
提出UniBEVFusion网络，通过引入RDL模块融合雷达特有信息提升视觉BEV特征质量，并采用UFF方法共享模块跨模态提取特征，在TJ4D数据集上取得SOTA表现。
摘要：
4D毫米波（MMW）雷达相比3D MMW雷达提供了高度信息和密集点云数据，在3D目标检测中日益普及。近年来，雷达-视觉融合模型展现出接近基于LiDAR模型的性能，在硬件成本更低和极端条件下鲁棒性更好方面具有优势。然而，许多雷达-视觉融合模型将雷达视为稀疏的LiDAR，未能充分利用雷达特有信息。此外，这些多模态网络通常对单一模态（尤其是视觉）的失效很敏感。为了解决这些挑战，我们提出了Radar Depth Lift-Splat-Shoot (RDL) 模块，将雷达特有数据整合到深度预测过程中，提升了视觉鸟瞰图 (BEV) 特征的质量。我们进一步引入了统一特征融合 (UFF) 方法，使用共享模块跨不同模态提取BEV特征。为了评估多模态模型的鲁棒性，我们开发了一种新颖的失效测试 (FT) 消融实验，通过注入高斯噪声来模拟视觉模态失效。我们在View-of-Delft (VoD) 和TJ4D数据集上进行了大量实验。结果表明，我们提出的统一BEVFusion (UniBEVFusion) 网络在TJ4D数据集上显著优于现有最先进模型，3D和BEV目标检测精度分别提升了1.44和1.72。
英文摘要：
4D millimeter-wave (MMW) radar, which provides both height information and dense point cloud data over 3D MMW radar, has become increasingly popular in 3D object detection. In recent years, radar-vision fusion models have demonstrated performance close to that of LiDAR-based models, offering advantages in terms of lower hardware costs and better resilience in extreme conditions. However, many radar-vision fusion models treat radar as a sparse LiDAR, underutilizing radar-specific information. Additionally, these multi-modal networks are often sensitive to the failure of a single modality, particularly vision. To address these challenges, we propose the Radar Depth Lift-Splat-Shoot (RDL) module, which integrates radar-specific data into the depth prediction process, enhancing the quality of visual Bird-Eye View (BEV) features. We further introduce a Unified Feature Fusion (UFF) approach that extracts BEV features across different modalities using shared module. To assess the robustness of multi-modal models, we develop a novel Failure Test (FT) ablation experiment, which simulates vision modality failure by injecting Gaussian noise. We conduct extensive experiments on the View-of-Delft (VoD) and TJ4D datasets. The results demonstrate that our proposed Unified BEVFusion (UniBEVFusion) network significantly outperforms state-of-the-art models on the TJ4D dataset, with improvements of 1.44 in 3D and 1.72 in BEV object detection accuracy.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
