arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2209.01745cs.CV
收藏

SEFormer:用于三维目标检测的结构嵌入式Transformer

SEFormer: Structure Embedding Transformer for 3D Object Detection

Tsinghua University(清华大学);Australian National University(澳大利亚国立大学);National University of Singapore(新加坡国立大学)

Xiaoyu Feng, Heming Du, Yueqi Duan, Yongpan Liu, Hehe Fan

摘要:

在不规则稀疏的LiDAR点云数据中有效地保留和编码目标的结构特征是点云数据上三维目标检测的关键挑战。最近,Transformer在许多2D甚至3D视觉任务中表现出了良好的性能。与固定和刚性卷积核相比,Transformer中的自注意力机制可以自适应地排除不相关或噪声点,因此适用于保持不规则LiDAR点云的局部空间结构。然而,Transformer仅基于自我注意力机制对点要素执行简单的求和,并且所有点共享相同的值变换。这种各向同性操作缺乏捕获对于3D目标检测重要的方向-距离-定向的局部结构的能力。本文提出了一种结构嵌入transFormer(SEFormer),它既能像传统的Transformer一样保持局部结构,又能对局部结构进行编码。与传统Transformer中的自注意力机制相比,SEFormer基于到查询点的相对方向和距离来学习值点的不同特征转换。然后,我们提出了一个SEFormer为基础的网络高性能的三维目标检测。大量的实验表明,该架构可以在Waymo Open Dataset上实现SOTA结果,Waymo Open Dataset是自动驾驶领域最大的3D检测基准。具体而言,SEFormer实现了79.02%的mAP,比现有工程高出1.2%。我们会发布密码的。

英文摘要:

Effectively preserving and encoding structure features from objects in irregular and sparse LiDAR points is a key challenge to 3D object detection on point cloud. Recently, Transformer has demonstrated promising performance on many 2D and even 3D vision tasks. Compared with the fixed and rigid convolution kernels, the self-attention mechanism in Transformer can adaptively exclude the unrelated or noisy points and thus suitable for preserving the local spatial structure in irregular LiDAR point cloud. However, Transformer only performs a simple sum on the point features, based on the self-attention mechanism, and all the points share the same transformation for value. Such isotropic operation lacks the ability to capture the direction-distance-oriented local structure which is important for 3D object detection. In this work, we propose a Structure-Embedding transFormer (SEFormer), which can not only preserve local structure as traditional Transformer but also have the ability to encode the local structure. Compared to the self-attention mechanism in traditional Transformer, SEFormer learns different feature transformations for value points based on the relative directions and distances to the query point. Then we propose a SEFormer based network for high-performance 3D object detection. Extensive experiments show that the proposed architecture can achieve SOTA results on Waymo Open Dataset, the largest 3D detection benchmark for autonomous driving. Specifically, SEFormer achieves 79.02% mAP, which is 1.2% higher than existing works. We will release the codes.

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。

↑