# Dual-Agent Reinforcement Learning for Adaptive and Cost-Aware Visual-Inertial Odometry

arXiv: https://arxiv.org/abs/2511.21083 · HTML: https://mlsanity.com/paper/2511.21083/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in RL for robot navigation and UAV control, one of 15 directions in Deep reinforcement learning methods.
- Builds on 24 of its 43 references that are in the map.
- First posted on arXiv 26 Nov 2025.

## Placed in

- Deep reinforcement learning methods / [RL for robot navigation and UAV control](https://mlsanity.com/deeprl/rl-for-robot-navigation-and-uav-control/)

## Builds on (most cited first, 24 of 24)

- [ORB-SLAM3: An Accurate Open-Source Library for Visual, Visual-Inertial and Multi-Map SLAM](https://mlsanity.com/paper/2007.11898/) (2020)
- [DROID-SLAM: Deep Visual SLAM for Monocular, Stereo, and RGB-D Cameras](https://mlsanity.com/paper/2108.10869/) (2021)
- [TartanAir: A Dataset to Push the Limits of Visual SLAM](https://mlsanity.com/paper/2003.14338/) (2020)
- [VINS-Mono: A Robust and Versatile Monocular Visual-Inertial State Estimator](https://mlsanity.com/paper/1708.03852/) (2017)
- [Learning to Explore using Active Neural SLAM](https://mlsanity.com/paper/2004.05155/) (2020)
- [Deep Patch Visual Odometry](https://mlsanity.com/paper/2208.04726/) (2022)
- [CodeSLAM - Learning a Compact, Optimisable Representation for Dense Visual SLAM](https://mlsanity.com/paper/1804.00874/) (2018)
- [The TUM VI Benchmark for Evaluating Visual-Inertial Odometry](https://mlsanity.com/paper/1804.06120/) (2018)
- [IONet: Learning to Cure the Curse of Drift in Inertial Odometry](https://mlsanity.com/paper/1802.02209/) (2018)
- [A General Optimization-based Framework for Global Pose Estimation with Multiple Sensors](https://mlsanity.com/paper/1901.03642/) (2019)
- [Direct Sparse Visual-Inertial Odometry using Dynamic Marginalization](https://mlsanity.com/paper/1804.05625/) (2018)
- [Goal-Driven Autonomous Exploration Through Deep Reinforcement Learning](https://mlsanity.com/paper/2103.07119/) (2021)
- [DeepVIO: Self-supervised Deep Learning of Monocular Visual Inertial Odometry using 3D Geometric Constraints](https://mlsanity.com/paper/1906.11435/) (2019)
- [SelfVIO: Self-Supervised Deep Monocular Visual-Inertial Odometry and Depth Estimation](https://mlsanity.com/paper/1911.09968/) (2019)
- [DM-VIO: Delayed Marginalization Visual-Inertial Odometry](https://mlsanity.com/paper/2201.04114/) (2022)
- [Deep Learning for Inertial Positioning: A Survey](https://mlsanity.com/paper/2303.03757/) (2023)
- [Efficient Deep Visual and Inertial Odometry with Adaptive Visual Modality Selection](https://mlsanity.com/paper/2205.06187/) (2022)
- [Deep Reinforcement Learning-based Large-scale Robot Exploration](https://mlsanity.com/paper/2403.10833/) (2024)
- [iSLAM: Imperative SLAM](https://mlsanity.com/paper/2306.07894/) (2023)
- [Reinforcement Learning Meets Visual Odometry](https://mlsanity.com/paper/2407.15626/) (2024)
- [CoVIO: Online Continual Learning for Visual-Inertial Odometry](https://mlsanity.com/paper/2303.10149/) (2023)
- [DBA-Fusion: Tightly Integrating Deep Dense Visual Bundle Adjustment with Multiple Sensors for Large-Scale Localization and Mapping](https://mlsanity.com/paper/2403.13714/) (2024)
- [Adaptive VIO: Deep Visual-Inertial Odometry with Online Continual Learning](https://mlsanity.com/paper/2405.16754/) (2024)
- [MoE-Gyro: Self-Supervised Over-Range Reconstruction and Denoising for MEMS Gyroscopes](https://mlsanity.com/paper/2506.06318/) (2025)

## Abstract (arXiv)

Visual-Inertial Odometry (VIO) is a critical component for robust ego-motion estimation, enabling foundational capabilities such as autonomous navigation in robotics and real-time 6-DoF tracking for augmented reality. Existing methods face a well-known trade-off: filter-based approaches are efficient but prone to drift, while optimization-based methods, though accurate, rely on computationally prohibitive Visual-Inertial Bundle Adjustment (VIBA) that is difficult to run on resource-constrained platforms. Rather than removing VIBA altogether, we aim to reduce how often and how heavily it must be invoked. To this end, we cast two key design choices in modern VIO, when to run the visual frontend and how strongly to trust its output, as sequential decision problems, and solve them with lightweight reinforcement learning (RL) agents. Our framework introduces a lightweight, dual-pronged RL policy that serves as our core contribution: (1) a Select Agent intelligently gates the entire VO pipeline based only on high-frequency IMU data; and (2) a composite Fusion Agent that first estimates a robust velocity state via a supervised network, before an RL policy adaptively fuses the full (p, v, q) state. Experiments on the EuRoC MAV and TUM-VI datasets show that, in our unified evaluation, the proposed method achieves a more favorable accuracy-efficiency-memory trade-off than prior GPU-based VO/VIO systems: it attains the best average ATE while running up to 1.77 times faster and using less GPU memory. Compared to classical optimization-based VIO systems, our approach maintains competitive trajectory accuracy while substantially reducing computational load.