aiqbench analysis
This research introduces a novel attention-based architecture that tokenizes inertial sensor data from legged robots to dynamically weigh measurements based on foot contact conditions. By eliminating reliance on explicit contact estimators, the method demonstrates superior proprioceptive state estimation in unseen terrains and gait patterns compared to existing learning-based or model-based approaches. The work advances robotics through a machine learning framework specifically designed for handling complex kinematic uncertainties without stationary assumptions.
GAIT: Legged Robot Proprioceptive State Estimation with Attention over Inertial-Leg Tokens
arXiv:2606.14160v1 Announce Type: new Abstract: In this paper, we propose a method that applies Inertial-Leg (IL) tokenization to an attention-based network for proprioceptive state estimation in legged robots. Unlike existing learning-based state estimators that concatenate all sensor measurements into a single flat vector, the proposed architecture represents inertial measurements and leg-wise measurements as individual tokens and uses an attention mechanism to learn the relative importance of each measurement.This design allows the network to reweight each measurement according to the current contact condition, reflecting the fact that the reliability of forward kinematic measurements depends on whether the corresponding foot is in contact. Unlike conventional contact-aided estimators, however, the proposed method learns this behavior without relying on an explicit contact estimator or on explicit measurement updates based on a stationary contact assumption. To validate the proposed method, we conducted experiments on a Unitree Go1 robot, including debris terrain not modeled in simulation and gait patterns not seen during training. Experimental results show that the proposed method achieves better estimation performance than existing learning-based state estimators under unseen gait patterns and also improves performance over contact-aided model-based methods.
Abstract:In this paper, we propose a method that applies Inertial-Leg (IL) tokenization to an attention-based network for proprioceptive state estimation in legged robots. Unlike existing learning-based state estimators that concatenate all sensor measurements into a single flat vector, the proposed architecture represents inertial measurements and leg-wise measurements as individual tokens and uses an attention mechanism to learn the relative importance of each this http URL design allows the network to reweight each measurement according to the current contact condition, reflecting the fact that the reliability of forward kinematic measurements depends on whether the corresponding foot is in contact. Unlike conventional contact-aided estimators, however, the proposed method learns this behavior without relying on an explicit contact estimator or on explicit measurement updates based on a stationary contact assumption. To validate the proposed method, we conducted experiments on a Unitree Go1 robot, including debris terrain not modeled in simulation and gait patterns not seen during training. Experimental results show that the proposed method achieves better estimation performance than existing learning-based state estimators under unseen gait patterns and also improves performance over contact-aided model-based methods.
Robotics (cs.RO)
arXiv:2606.14160 [cs.RO]
(or arXiv:2606.14160v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2606.14160 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 · 275 words
Related coverage
KILVO: Kinematic-Inertial-LiDAR-Visual Odometry with Robust Multimodal Adaptation for Humanoid Robots
arXiv · Aug 7, 2026
Interacting Multiple Model Proprioceptive Odometry for Legged Robots
arXiv · Jul 22, 2026
DogLegs: Robust Proprioceptive State Estimation for Legged Robots Using Multiple Leg-Mounted IMUs
arXiv · Sep 4, 2026
LAG-IO: Multi-IMU State Estimation for Legged Robots Using Factor Graphs
arXiv · Oct 14, 2026