aiqbench analysis
This research addresses a specific localization challenge in visual-inertial SLAM by decoupling robot motion from the vertical transport of elevators, enabling robust mapping during transit. By treating elevator movement as a distinct state variable within bundle adjustment and leveraging vision-language models for zero-shot detection, the system maintains high-fidelity tracking without requiring manual intervention. The work advances the practical deployment of 3D Gaussian Splatting in dynamic environments where standard visual-inertial estimators typically fail due to coordinate frame conflicts.
Elevator-VIGS: Separating Elevator Motion from Robot Motion in Visual-Inertial Gaussian Splatting SLAM
arXiv:2609.23491v1 Announce Type: new Abstract: We present Elevator-VIGS, a visual-inertial 3D Gaussian Splatting SLAM system that keeps tracking and mapping through elevator rides. Inside a moving elevator, the two sensors are in conflict. The camera sees only the robot's motion relative to the elevator, while the IMU senses that motion plus the elevator's motion relative to the world. This conflict is challenging for existing visual-inertial estimators. If vision dominates, the estimator tracks only the robot's motion within the elevator and misses the elevator's rise, and if the conflict remains, the estimator diverges. We observe that the conflict comes from forcing both observations into a single coordinate frame. We instead estimate the robot's pose in the elevator's coordinate frame, and the elevator's motion relative to the world as a per-keyframe transport state, the elevator's rise and vertical velocity, within dense visual-inertial bundle adjustment. Elevator-VIGS detects rides zero-shot with a vision-language model and a depth network, and constrains the transport state at the departure and the arrival. We record real-world and simulated elevator sequences. On these sequences, Elevator-VIGS achieves state-of-the-art tracking and rendering performance. On four elevator-free public benchmarks it keeps the state-of-the-art performance of VIGS-SLAM. Project page: https://ruizhou-cn.github.io/elevator-vigs/.
Abstract:We present Elevator-VIGS, a visual-inertial 3D Gaussian Splatting SLAM system that keeps tracking and mapping through elevator rides. Inside a moving elevator, the two sensors are in conflict. The camera sees only the robot's motion relative to the elevator, while the IMU senses that motion plus the elevator's motion relative to the world. This conflict is challenging for existing visual-inertial estimators. If vision dominates, the estimator tracks only the robot's motion within the elevator and misses the elevator's rise, and if the conflict remains, the estimator diverges. We observe that the conflict comes from forcing both observations into a single coordinate frame. We instead estimate the robot's pose in the elevator's coordinate frame, and the elevator's motion relative to the world as a per-keyframe transport state, the elevator's rise and vertical velocity, within dense visual-inertial bundle adjustment. Elevator-VIGS detects rides zero-shot with a vision-language model and a depth network, and constrains the transport state at the departure and the arrival. We record real-world and simulated elevator sequences. On these sequences, Elevator-VIGS achieves state-of-the-art tracking and rendering performance. On four elevator-free public benchmarks it keeps the state-of-the-art performance of VIGS-SLAM. Project page: this https URL.
Robotics (cs.RO)
arXiv:2609.23491 [cs.RO]
(or arXiv:2609.23491v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2609.23491 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 287 words
Related coverage
Make a Soft Digital Clock Tick With Millifluidics
IEEE Spectrum · May 29, 2026
Dyna-2: General Purpose Physical Agents with Hour-Long Autonomy — Jason Ma, Dyna Robotics
Latent Space · Oct 13, 2026
Uber is building an autonomous vehicle empire, and here’s every company it’s using to do it
TechCrunch · Aug 1, 2026
From Target Selection to Digging: A Learning-Based Framework for Continuous Autonomous Excavation
arXiv · Sep 25, 2026