Accelerating SfM-based Pose Estimation with Dominating Set

2025-06-04

AI Summary

To address the high computational cost and latency limitations of Structure-from-Motion (SfM) pose estimation in real-time applications such as AR/VR and robotics, this work pioneers the integration of graph-theoretic dominating set theory into SfM model compression. We propose a collaborative preprocessing framework based on a feature-association graph, which jointly prunes reference images and 3D points via dominating set construction while preserving topological consistency and substantially reducing optimization dimensionality. The method comprises three stages: graph construction, dominating-set-driven pruning, and multi-scale pose refinement. Evaluated on the OnePose dataset, our approach achieves 1.5–14.48× speedup in inference time, reduces the number of reference images by 17–23×, compresses the point cloud size by 2.27–4×, and incurs negligible accuracy degradation—pose errors increase by less than 1.2°/cm.

Technology Category

• Computer Vision: Other Foundations of Computer Vision
• Machine Learning: Learning on the Edge & Model Compression
• Intelligent Robots: State Estimation

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines

Abstract

This paper introduces a preprocessing technique to speed up Structure-from-Motion (SfM) based pose estimation, which is critical for real-time applications like augmented reality (AR), virtual reality (VR), and robotics. Our method leverages the concept of a dominating set from graph theory to preprocess SfM models, significantly enhancing the speed of the pose estimation process without losing significant accuracy. Using the OnePose dataset, we evaluated our method across various SfM-based pose estimation techniques. The results demonstrate substantial improvements in processing speed, ranging from 1.5 to 14.48 times, and a reduction in reference images and point cloud size by factors of 17-23 and 2.27-4, respectively. This work offers a promising solution for efficient and accurate 3D pose estimation, balancing speed and accuracy in real-time applications.

Problem

• Speeding up SfM-based pose estimation for real-time applications
• Reducing processing time without significant accuracy loss
• Balancing speed and accuracy in AR, VR, and robotics

Innovation

• Dominating set from graph theory
• Preprocesses SfM models efficiently
• Balances speed and accuracy