English

Boosting the Performance of Object Tracking with a Half-Precision Particle Filter on GPU

Distributed, Parallel, and Cluster Computing 2023-08-03 v1

Abstract

High-performance GPU-accelerated particle filter methods are critical for object detection applications, ranging from autonomous driving, robot localization, to time-series prediction. In this work, we investigate the design, development and optimization of particle-filter using half-precision on CUDA cores and compare their performance and accuracy with single- and double-precision baselines on Nvidia V100, A100, A40 and T4 GPUs. To mitigate numerical instability and precision losses, we introduce algorithmic changes in the particle filters. Using half-precision leads to a performance improvement of 1.5-2x and 2.5-4.6x with respect to single- and double-precision baselines respectively, at the cost of a relatively small loss of accuracy.

Keywords

Cite

@article{arxiv.2308.00763,
  title  = {Boosting the Performance of Object Tracking with a Half-Precision Particle Filter on GPU},
  author = {Gabin Schieffer and Nattawat Pornthisan and Daniel Araújo de Medeiros and Stefano Markidis and Jacob Wahlgren and Ivy Peng},
  journal= {arXiv preprint arXiv:2308.00763},
  year   = {2023}
}

Comments

12 pages, 8 figures, conference. To be published in The 21st International Workshop on Algorithms, Models and Tools for Parallel Computing on Heterogeneous Platforms (HeteroPar2023)

R2 v1 2026-06-28T11:45:52.653Z