Inertial navigation using low-cost MEMS sensors is plagued by rapid drift due to sensor noise and bias instability. While recent data-driven approaches have made significant strides, they often struggle with micro-drifts during stationarity and mode fusion during complex motion transitions due to their reliance on fixed-window regression. In this work, we introduce Gated Neural Inertial Odometry (GNIO), a novel learning-based framework that explicitly models motion validity and context. We propose two key architectural innovations: \ding{182} a learnable Motion Bank that queries a global dictionary of motion patterns to provide semantic context beyond the local receptive field, and \ding{183} a Gated Prediction Head that decomposes displacement into magnitude and direction. This gating mechanism acts as a soft, differentiable Zero-Velocity Update (ZUPT), dynamically suppressing sensor noise during stationary periods while scaling predictions during dynamic motion. Extensive experiments across four public benchmarks demonstrate that GNIO significantly reduces position drift compared to state-of-the-art CNN and Transformer-based baselines. Notably, GNIO achieves a 60.21% reduction in trajectory error on the OxIOD dataset and exhibits superior generalization in challenging scenarios involving frequent stops and irregular motion speeds.
Figures & tables
Figure 1: System Overview of GNIO.
Architecture
Model Family
Params (M)
GFLOPs
CNN-based
TLIO
ResNet-18+EKF
5.42
0.08
DeepILS
Lightweight ResNet
2.29
0.03
GNIO (Ours)
Memory + Gated
4.90
0.08
Transformer-based
CTIN
Contextual Transformer
0.56
7.27
Table I: Comparison of model parameters and computational complexity.
Methods
OxIOD [ 21 ]
RIDI [ 15 ]
RoNIN [ 2 ]
IDOL [ 22 ]
seen
unseen
seen
unseen
seen
unseen
seen
unseen
TLIO [ 3 ]
2.67
2.54
1.58
1.49
5.95
7.10
5. 89
5.34
CTIN [ 12 ]
2.32
3.34
1.39
1.86
4.62
5.61
2.90
3.69
iMoT [ 13 ]
1.86
0.90
1.68
1.49
3.78
5.31
2.22
3.00
DeepILS [ 23 ]
–
–
1.56
1.48
4.80
6.80
–
–
GNIO (Ours)
0.74↑60.21%
0.71↑21.11%
1.22↑12.23%
1.19↑19.59%
4.31↓14.02%
5.02↑5.46%
3.45↓55.45%
3.89↓29.66%
Table II: Trajectory Prediction Comparision of GNIO and Baseline Models.
Configuration
Motion Bank
Gated Prediction Head
RMSE ( m )
Gain ( % )
Baseline (TLIO)
–
–
2.40
–
Context Enriched
✓
–
1.98
17.5
Structurally Gated
–
✓
1.84
23.3
GNIO (Full)
✓
✓
1.66
30.8
Table III: Component-wise Ablation Analysis on TLIO Dataset.
Gating σ(⋅)
Scaling ψ(⋅)
Scale Formula
RMSE ( m )
Gating Activation Analysis
Sigmoid
Linear
x
2.05
Tanh
Linear
x
2.00
Magnitude Mapping Analysis (Fixed Tanh Gating)
Tanh
Exp
ex
1.94
Tanh
Abs
∣x∣
1.92
Table IV: Design Ablation of the Gated Prediction Head on the TLIO Dataset.
Figure 2: Impact of Motion Bank Size on Trajectory Error
Figure 3: Qualitative comparison of estimated trajectories.
Neural inertial odometry increasingly uses networks as learned measurements inside filtering pipelines. Such measurements should transform consistently under arbitrary IMU mounting conventions: their mean must transform as a vector, and their covariance must transform congruently as a second-order tensor. We present GINIO, a geometric SO(3)-equivariant interface for neural inertial odometry under arbitrary rotations of the IMU measurement frame. Given calibrated IMU windows, our framework predicts a motion measurement and uncertainty obeying these tensorial laws. To support efficient sensor-frame learning, we introduce Last-Frame Alignment (LFA), a deterministic preprocessing step that is provably equivalent to world-frame training for SO(3)-equivariant predictors. The connected estimator tracks sensor-local states such as IMU bias, separating nuisance estimation from the geometric law enforced by the learned measurement. We instantiate the same interface in filter-connected NIO, AirIO-style recurrent aerial prediction, EqNIO-style full-SO(3) canonicalization, and ResNet-style temporal backbones. On TLIO, GINIO achieves 2.018 m ID/SO(3) ATE while EqNIO degrades to 76.389 m, using 11.6x fewer FLOPs. On NanoBench, our AirIO-style instantiation improves ATE from 5.579 m to 1.430 m without external attitude input, and our ResNet-style instantiation reaches 0.581 m ATE versus 0.645 m for ResNet1D. On Fetch, GINIO empirically reduces unseen physical-remount ATE from 8.15 m to 0.50 m without retraining, demonstrating robustness beyond the exact coordinate-frame guarantee. For uncertainty, spectral covariance reduces covariance-equivariance error by over three orders of magnitude compared with a diagonal head.
Chankyo Kim, Minghan Zhu, Tzu-Yuan Lin +2
University of Michigan Ann Arbor, MI, United States
The accuracy of consumer-grade inertial navigation is bottlenecked by the stochastic noise of Micro-Electro-Mechanical Systems (MEMS). Traditional deterministic neural architectures often succumb to ``estimation jittering,'' sacrificing high-frequency kinematic fidelity for numerical stability. We propose PedestrianDiffusion, a multimodal spectral-domain generative framework reformulating dense 6D state estimation as a continuous conditional denoising process. By operating in the frequency domain, our formulation bounds the spectral covariance, acting as a mathematical preconditioner to stabilize the reverse diffusion trajectory. Furthermore, we introduce a zero-shot semantic conditioning mechanism leveraging vision-language embeddings as categorical priors to generalize across heterogeneous sensor noise profiles. To address the computational intractability of generative tracking, we deploy a single-step deterministic probability flow ODE solver (T=1). This yields high-capacity asynchronous batch trajectory refinement, establishing the viability of generative architectures for asynchronous batch trajectory refinement on edge hardware. Extensive evaluations on the OxIOD, RIDI, RoNIN, and TLIO benchmarks demonstrate that PedestrianDiffusion achieves state-of-the-art performance, exhibiting unprecedented robustness to impulse perturbations and coupled 6D kinematic drift. This work provides a rigorous algorithmic blueprint for next-generation Neural Inertial Measurement Units (N-IMUs).
Neural inertial odometry has demonstrated strong potential for motion estimation in challenging environments, yet inertial-only preintegration remains sensitive to IMU bias and uncertainty. To this end, this paper introduces PLATO:Preintegration Learning from Accurate Trajectory Observations, a likelihood-based framework that leverages accurate trajectory observations to jointly learn IMU bias dynamics modeled by a neural ordinary differential equation(NODE) and gyroscope and accelerometer noise covariances. Optimization exploits the sparse structure of the negative log-likelihood, with IMU noise-parameter gradients computed by forward differentiation. A tailored double-adjoint scheme couples a discrete invariant-error adjoint with a continuous-time adjoint for the bias NODE, enabling memory-efficient likelihood optimization over the nested bias-dynamics and IMU-preintegration rollouts. Validation on EuRoC shows improved performance, and underwater robot experiments demonstrate applicability under intermittent lighting failures and visual degradation.