BEVDetNet: Bird's Eye View LiDAR Point Cloud based Real-time 3D Object Detection for Autonomous Driving

TL;DR

This paper introduces BEVDetNet, a real-time LiDAR-based 3D object detection model optimized for embedded systems, achieving low latency and high speed with minimal accuracy loss for autonomous driving.

Contribution

It presents a novel BEV-based semantic segmentation architecture that is highly efficient, deterministic, and extendable, suitable for embedded platforms in autonomous vehicles.

Findings

4 ms latency on Nvidia Xavier platform

5x faster than top accuracy models

2% minimal accuracy degradation

Abstract

3D object detection based on LiDAR point clouds is a crucial module in autonomous driving particularly for long range sensing. Most of the research is focused on achieving higher accuracy and these models are not optimized for deployment on embedded systems from the perspective of latency and power efficiency. For high speed driving scenarios, latency is a crucial parameter as it provides more time to react to dangerous situations. Typically a voxel or point-cloud based 3D convolution approach is utilized for this module. Firstly, they are inefficient on embedded platforms as they are not suitable for efficient parallelization. Secondly, they have a variable runtime due to level of sparsity of the scene which is against the determinism needed in a safety system. In this work, we aim to develop a very low latency algorithm with fixed runtime. We propose a novel semantic segmentation architecture as a single unified model for object center detection using key points, box predictions and orientation prediction using binned classification in a simpler Bird's Eye View (BEV) 2D representation. The proposed architecture can be trivially extended to include semantic segmentation classes like road without any additional computation. The proposed model has a latency of 4 ms on the embedded Nvidia Xavier platform. The model is 5X faster than other top accuracy models with a minimal accuracy degradation of 2% in Average Precision at IoU=0.5 on KITTI dataset.