TRACE: Learned Proprioceptive Odometry for Legged Robots under Unreliable Contact Conditions

Published August 7, 2026 · Category: Robotics

Overview

arXiv:2608.05975v1 Announce Type: new Abstract: In this paper, we present TRACE (Tokenized Robust Attention for Contact-Aware Estimation), an end-to-end learned proprioceptive odometry estimator for legged robots under unreliable contact conditions. The proposed estimator directly predicts relative displacement, relative rotation, and body-frame velocity from a recent history of onboard inertial and joint measurements. To improve robustness under unreliable contact conditions, we introduce a foot-aware cross-attention module that adaptively weights IMU and leg-wise kinematic tokens without relying on manually defined contact or slip thresholds. The estimator is trained with direct supervision and two physics-inspired auxiliary losses that promote kinematic consistency and reliable use of leg information. To reduce policy-specific overfitting and consequently improve sim-to-real transfer, simulation training incorporates policy randomization, followed by partial real-world fine-tuning of the temporal encoder and prediction head. Experiments across diverse indoor and outdoor terrains demonstrate consistent reductions in position drift compared with classical filtering-based, hybrid, and purely learning-based baselines. Ablation studies further validate the contributions of the proposed training objectives, policy randomization, and real-world fine-tuning, particularly under unreliable contacts and sim-to-real mismatch.

Source

Originally published at arxiv.org.

Related Articles

• CADRE: Dynamic Catching via Implicit Contact Descriptors and Task-Appropriate Recovery Affordances
• Multi-Robot Motions in Milliseconds: Vector-Accelerated Primitives for Sampling-Based Planning
• Dream-MPC: Gradient-Based Model Predictive Control with Latent Imagination