Saved in:
Bibliographic Details
Main Authors: Wan, Ziyu, Zhao, Lin
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2507.15716
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908756372094976
author Wan, Ziyu
Zhao, Lin
author_facet Wan, Ziyu
Zhao, Lin
contents This paper proposes DiffPF, a differentiable particle filter that leverages diffusion models for state estimation in dynamic systems. Unlike conventional differentiable particle filters, which require importance weighting and typically rely on predefined or low-capacity proposal distributions. DiffPF learns a flexible posterior sampler by conditioning a diffusion model on predicted particles and the current observation. This enables accurate, equally-weighted sampling from complex, high-dimensional, and multimodal filtering distributions. We evaluate DiffPF across a range of scenarios, including both unimodal and highly multimodal distributions, and test it on simulated as well as real-world tasks, where it consistently outperforms existing filtering baselines. In particular, DiffPF achieves an 82.8% improvement in estimation accuracy on a highly multimodal global localization benchmark, and a 26% improvement on the real-world KITTI visual odometry benchmark, compared to state-of-the-art differentiable filters. To the best of our knowledge, DiffPF is the first method to integrate conditional diffusion models into particle filtering, enabling high-quality posterior sampling that produces more informative particles and significantly improves state estimation.
format Preprint
id arxiv_https___arxiv_org_abs_2507_15716
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle DiffPF: Differentiable Particle Filtering with Generative Sampling via Conditional Diffusion Models
Wan, Ziyu
Zhao, Lin
Robotics
This paper proposes DiffPF, a differentiable particle filter that leverages diffusion models for state estimation in dynamic systems. Unlike conventional differentiable particle filters, which require importance weighting and typically rely on predefined or low-capacity proposal distributions. DiffPF learns a flexible posterior sampler by conditioning a diffusion model on predicted particles and the current observation. This enables accurate, equally-weighted sampling from complex, high-dimensional, and multimodal filtering distributions. We evaluate DiffPF across a range of scenarios, including both unimodal and highly multimodal distributions, and test it on simulated as well as real-world tasks, where it consistently outperforms existing filtering baselines. In particular, DiffPF achieves an 82.8% improvement in estimation accuracy on a highly multimodal global localization benchmark, and a 26% improvement on the real-world KITTI visual odometry benchmark, compared to state-of-the-art differentiable filters. To the best of our knowledge, DiffPF is the first method to integrate conditional diffusion models into particle filtering, enabling high-quality posterior sampling that produces more informative particles and significantly improves state estimation.
title DiffPF: Differentiable Particle Filtering with Generative Sampling via Conditional Diffusion Models
topic Robotics
url https://arxiv.org/abs/2507.15716