AFDet: Anchor Free One Stage 3D Object Detection

TL;DR

AFDet introduces an anchor-free, one-stage 3D object detection method optimized for embedded systems, simplifying post-processing and tuning while maintaining competitive accuracy on major datasets.

Contribution

This paper presents the first anchor-free, NMS-free one-stage 3D detection framework, reducing complexity and computational cost compared to anchor-based methods.

Findings

Performs competitively on KITTI and Waymo datasets

Simplifies post-processing and anchor tuning

Efficiently runs on CNN accelerators or GPUs

Abstract

High-efficiency point cloud 3D object detection operated on embedded systems is important for many robotics applications including autonomous driving. Most previous works try to solve it using anchor-based detection methods which come with two drawbacks: post-processing is relatively complex and computationally expensive; tuning anchor parameters is tricky. We are the first to address these drawbacks with an anchor free and Non-Maximum Suppression free one stage detector called AFDet. The entire AFDet can be processed efficiently on a CNN accelerator or a GPU with the simplified post-processing. Without bells and whistles, our proposed AFDet performs competitively with other one stage anchor-based methods on KITTI validation set and Waymo Open Dataset validation set.