ViSTA-SLAM: Visual SLAM with Symmetric Two-view Association

2025-09-01

AI Summary

Existing monocular SLAM systems suffer from dependence on calibrated intrinsic parameters, architectural complexity, and low-quality two-view geometric constraints. To address these issues, this paper proposes a lightweight, real-time, intrinsic-free monocular SLAM framework. Methodologically, the front-end employs a symmetric dual-view association network that jointly estimates relative pose and local point clouds directly from two RGB frames; the back-end performs Sim(3) pose-graph optimization with loop closure to achieve globally consistent, scale-aware mapping. The proposed network has only 35% of the parameters of state-of-the-art methods, significantly improving robustness of two-view constraints and cross-camera generalization. Extensive experiments on multiple standard benchmarks demonstrate superior trajectory accuracy and dense reconstruction quality compared to current best methods, while maintaining real-time performance.

Technology Category

• Computer Vision: Learning & Optimization for CV
• Intelligent Robots: State Estimation
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

We present ViSTA-SLAM as a real-time monocular visual SLAM system that operates without requiring camera intrinsics, making it broadly applicable across diverse camera setups. At its core, the system employs a lightweight symmetric two-view association (STA) model as the frontend, which simultaneously estimates relative camera poses and regresses local pointmaps from only two RGB images. This design reduces model complexity significantly, the size of our frontend is only 35% that of comparable state-of-the-art methods, while enhancing the quality of two-view constraints used in the pipeline. In the backend, we construct a specially designed Sim(3) pose graph that incorporates loop closures to address accumulated drift. Extensive experiments demonstrate that our approach achieves superior performance in both camera tracking and dense 3D reconstruction quality compared to current methods. Github repository: https://github.com/zhangganlin/vista-slam

Problem

• Monocular visual SLAM without camera intrinsics requirement
• Lightweight symmetric two-view association for pose estimation
• Sim(3) pose graph with loop closures for drift correction

Innovation

• Monocular SLAM without camera intrinsics
• Lightweight symmetric two-view association frontend
• Sim(3) pose graph with loop closure backend