改进旋转位置编码，提升纯摄像头3D检测的时序建模能力
RoPETR: Improving Temporal Camera-Only 3D Detection by Integrating Enhanced Rotary Position Embedding
- 设计定制化旋转位置编码增强时序建模
- 在NuScenes上达到70.86%的NDS新纪录
- 适合关注纯视觉3D检测时序性能的研究者
本技术报告针对StreamPETR框架在纯摄像头3D目标检测中速度估计能力不足的问题提出改进。尽管该框架在边界框检测上表现优异（高mAP），但在NuScenes数据集上的分析表明速度估计是主要瓶颈。为此，我们设计了一种定制化的旋转位置编码策略，以增强模型的时序建模能力。在NuScenes测试集上的实验结果表明，采用ViT-L骨干网络时，所提方法取得了70.86%的NDS，刷新了纯摄像头3D检测的性能基准。
原文摘要 · Abstract (English)
This technical report introduces a targeted improvement to the StreamPETR framework, specifically aimed at enhancing velocity estimation, a critical factor influencing the overall NuScenes Detection Score. While StreamPETR exhibits strong 3D bounding box detection performance as reflected by its high mean Average Precision our analysis identified velocity estimation as a substantial bottleneck when evaluated on the NuScenes dataset. To overcome this limitation, we propose a customized positional embedding strategy tailored to enhance temporal modeling capabilities. Experimental evaluations conducted on the NuScenes test set demonstrate that our improved approach achieves a state-of-the-art NDS of 70.86% using the ViT-L backbone, setting a new benchmark for camera-only 3D object detection.
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。