Sparse PointPillars: Maintaining and Exploiting Input Sparsity to Improve Runtime on Embedded Systems

TL;DR

This paper introduces a BEV 3D object detection method that leverages input sparsity to significantly improve runtime efficiency on embedded systems, with minimal impact on detection accuracy.

Contribution

The authors develop a sparse BEV detection approach that maintains and exploits input sparsity, achieving faster runtimes on embedded hardware without sacrificing detection quality.

Findings

Achieves 2x or more speedup on embedded systems

Maintains detection performance with modest accuracy decrease

Demonstrates effectiveness on KITTI and Matterport-Chair datasets

Abstract

Bird's Eye View (BEV) is a popular representation for processing 3D point clouds, and by its nature is fundamentally sparse. Motivated by the computational limitations of mobile robot platforms, we create a fast, high-performance BEV 3D object detector that maintains and exploits this input sparsity to decrease runtimes over non-sparse baselines and avoids the tradeoff between pseudoimage area and runtime. We present results on KITTI, a canonical 3D detection dataset, and Matterport-Chair, a novel Matterport3D-derived chair detection dataset from scenes in real furnished homes. We evaluate runtime characteristics using a desktop GPU, an embedded ML accelerator, and a robot CPU, demonstrating that our method results in significant detection speedups (2X or more) for embedded systems with only a modest decrease in detection quality. Our work represents a new approach for practitioners to optimize models for embedded systems by maintaining and exploiting input sparsity throughout their entire pipeline to reduce runtime and resource usage while preserving detection performance.