arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2308.04556cs.CV
收藏

FocalFormer3D:聚焦困难实例的三维目标检测

FocalFormer3D : Focusing on Hard Instance for 3D Object Detection

The Chinese University of Hong Kong(香港中文大学);NVIDIA(英伟达);Caltech(加州理工学院)

Yilun Chen, Zhiding Yu, Yukang Chen, Shiyi Lan, Animashree Anandkumar, Jiaya Jia, Jose Alvarez

AI总结:

本文提出困难实例探测流程HIP及其三维检测实例FocalFormer3D,通过多阶段查询生成和框级transformer解码器聚焦挖掘困难目标,在nuScenes和Waymo上取得领先的检测与跟踪性能。

摘要:

三维目标检测中的假阴性(FN),例如对行人、车辆或其他障碍物的漏检,可能导致自动驾驶中的潜在危险情况。尽管这一问题至关重要,但在许多当前的三维检测方法中尚未得到充分研究。本文提出困难实例探测(Hard Instance Probing,HIP),这是一种通用流程,以多阶段方式识别假阴性(FN),并引导模型专注于挖掘困难实例。针对三维目标检测,我们将该方法实例化为FocalFormer3D,这是一种简单而有效的检测器,擅长挖掘困难目标并提高预测召回率。FocalFormer3D具有多阶段查询生成机制以发现困难目标,并采用框级transformer解码器从大量目标候选中高效地区分目标。在nuScenes和Waymo数据集上的实验结果验证了FocalFormer3D的优越性能。这一优势在检测和跟踪任务上均带来了强劲表现,涵盖LiDAR和多模态设置。值得注意的是,FocalFormer3D在nuScenes检测基准上达到了70.5 mAP和73.9 NDS,在nuScenes跟踪基准上达到了72.1 AMOTA,两项均在nuScenes LiDAR排行榜上排名第一。我们的代码可在https://github.com/NVlabs/FocalFormer3D获取。

英文摘要:

False negatives (FN) in 3D object detection, {\em e.g.}, missing predictions of pedestrians, vehicles, or other obstacles, can lead to potentially dangerous situations in autonomous driving. While being fatal, this issue is understudied in many current 3D detection methods. In this work, we propose Hard Instance Probing (HIP), a general pipeline that identifies \textit{FN} in a multi-stage manner and guides the models to focus on excavating difficult instances. For 3D object detection, we instantiate this method as FocalFormer3D, a simple yet effective detector that excels at excavating difficult objects and improving prediction recall. FocalFormer3D features a multi-stage query generation to discover hard objects and a box-level transformer decoder to efficiently distinguish objects from massive object candidates. Experimental results on the nuScenes and Waymo datasets validate the superior performance of FocalFormer3D. The advantage leads to strong performance on both detection and tracking, in both LiDAR and multi-modal settings. Notably, FocalFormer3D achieves a 70.5 mAP and 73.9 NDS on nuScenes detection benchmark, while the nuScenes tracking benchmark shows 72.1 AMOTA, both ranking 1st place on the nuScenes LiDAR leaderboard. Our code is available at \url{https://github.com/NVlabs/FocalFormer3D}.

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。

↑