Pith. sign in

REVIEW 3 major objections 5 minor 41 references

Learning IMU Bias with Diffusion Model

T0 review · 3 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read IMU bias is a distribution, not a point; a diffusion model learns it and improves inertial-only odometry.

desk verdict First paper to treat IMU bias as a conditional distribution via diffusion, with a solid but modest IOO gain; the “faithful bias” claim is undermined by using VINS-recovered bias as both training target and evaluation reference without independent validation. read the letter →

arxiv 2505.11763 v1 pith:EI5BGJL4 submitted 2025-05-17 cs.RO

Shenghao Zhou , Saimouli Katragadda , Guoquan Huang This is my paper · ORCID
classification cs.RO
keywords IMUbiasconditionaldiffusionmodelinertial-onlyodometryprobabilisticmodelingdirectsupervisionEuRoCdatasetdeeplearningforinertialsensing
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper argues that IMU bias is inherently stochastic—it drifts with temperature, vibration, and power cycles—so predicting it as a single number throws away information. It therefore learns the conditional distribution $p(b_g,b_a\mid\omega_m,a_m)$ with a diffusion model and samples bias from that distribution for inertial-only odometry. On the EuRoC dataset, sampling from the diffusion model gives lower position error than AirIMU, the indirect-supervision state of the art, and than direct regression baselines, while bias traces match the reference bias in magnitude and smooth drift. The result matters because when a camera loses sight of the tracked object, IMU-only tracking has to carry the motion estimate, and better bias prediction is the main lever.

What carries the argument

The load-bearing object is the conditional denoiser $\epsilon_\theta(x_t,t,c)$: a reversed diffusion process that, given the noised bias $x_t$, the diffusion step $t$, and a condition code $c$, predicts the noise added at the previous step. The condition code is extracted from a one-second IMU window by a temporal convolutional network, and the denoiser backbone is a two-cell GRU followed by a linear layer, trained with the standard MSE noise-prediction loss. At inference the model uses DDIM to generate bias samples in 25 steps. The same backbone trained to regress a point estimate performs worse, which is what isolates the probabilistic formulation as the source of the improvement.

What would settle it

Use a dataset where IMU bias is measured independently of any estimator—for example, stationary sequences with controlled temperature cycling or a high-grade reference IMU—and compare diffusion samples with direct regression on bias fidelity and integrated position error; if the diffusion advantage disappears or samples no longer match the independent bias, the central claim fails.

Watch

Extended reading notes

Core claim

The central claim is that modeling the IMU-conditioned bias as a distribution rather than regressing a point estimate fixes a real failure mode of learned bias predictors. Indirect supervision through integrated motion can reward spurious corrections that are not the true bias; direct regression, even with true bias labels, cannot represent the uncertainty in the mapping from inertial readings to bias. A conditional diffusion model trained with direct bias supervision samples from this distribution, and the sampled bias is both closer to the recovered ground truth and more effective for integration. On the tested EuRoC sequences the method reaches an average position RMSE of 0.0475 m versus 0.0521 m for AirIMU, while keeping orientation error within a comparable range.

Load-bearing premise

The argument assumes that biases recovered from a visual-inertial joint optimization and interpolated to IMU rate are accurate enough to serve as ground truth for both training and evaluation; if those labels carry the estimator's errors, the diffusion model may learn to imitate the estimator rather than true sensor bias.

Editorial extensions

If this is right

  • Inertial-only odometry improves without assuming any particular motion pattern, because the bias model conditions only on IMU readings.
  • Direct supervision plus a distributional output avoids the spurious corrections that integration-based indirect supervision can learn.
  • The architecture is light enough for edge deployment: 2.2 million parameters and roughly 145 ms inference on an embedded GPU.
  • Because the output is a distribution, later work can fuse the full posterior into a filter or choose samples based on risk, not just take the mean.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the paper: the learned distribution's variance could be used as a process-noise or observation model in a Kalman filter, turning the bias samples into uncertainty-aware corrections rather than a single draw.
  • Beyond the paper: the same conditional-diffusion treatment should transfer to other unobservable, time-varying states in inertial sensing, such as scale errors, misalignment, or g-sensitivity, which the paper names but does not model.
  • Beyond the paper: training on biases pooled from several visual-inertial estimators might make the model more robust to any particular estimator's artifacts, but the paper tests only one recovery procedure.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper proposes to learn IMU bias with a conditional diffusion model, in contrast to prior regression-based methods. The bias is modeled as a probability distribution conditioned on IMU readings, with a TCN encoder extracting a condition code and a lightweight GRU-based denoiser predicting the diffusion noise. The training target is the bias recovered by VINS joint optimization, interpolated from frame rate to IMU rate. The authors evaluate inertial-only odometry (IOO) on four EuRoC sequences, reporting relative position RMSE and relative orientation error against AirIMU, direct regression baselines, a best-of-N random-walk baseline, and a U-Net diffusion variant. They claim that the probabilistic formulation yields better motion accuracy and more faithful bias predictions than regression-based approaches.

Significance. If the claims are fully validated, the paper makes a useful contribution: it applies conditional diffusion in a sensible way to IMU bias modeling, uses a lightweight architecture that is plausibly suitable for embedded deployment, and constructs a strong random-walk upper-bound baseline. The standard diffusion training objective is clearly described, and the comparison with both direct and indirect regression baselines is a reasonable experimental design. However, the central claims of improved accuracy and faithful bias prediction are not yet fully established: the ground-truth bias proxy is not independently validated, results are reported without error bars or statistical tests, and the bias-fidelity evidence is qualitative and anecdotal. With additional validation and statistical rigor, the contribution could be of interest to the inertial odometry and mobile robotics community; as it stands, the experimental support is insufficient for acceptance.

major comments (3)
  1. [Section IV-D] The ground-truth bias used for training and for the fidelity comparison in Fig. 2 is obtained from VINS joint optimization and interpolation, but its accuracy is never checked against an independent reference. Because the OpenVINS estimator includes a random-walk bias prior, the observation that the recovered bias is 'continuous and changes slowly' is exactly what the prior enforces and cannot independently validate the labels. The model is therefore trained and evaluated against a proxy that may inherit estimator artifacts such as smoothing, delay, and prior-induced smoothness. To support the 'faithful bias' claim, please validate the proxy against an independent source (for example, optimization with motion-capture measurements, stationary Allan-variance calibration, or synthetic IMU data with known bias), or explicitly downgrade the claim to 'consistent with VINS-recovered bias' and assess sensitivity by training with an alternative estimator.
  2. [Section V, Table I] Table I reports per-sequence and average PRMSE/ROE without error bars or significance testing, while the diffusion result is averaged over 50 samples and deterministic baselines are single runs. The headline gain is small (average PRMSE 0.0475 vs. 0.0521 for AirIMU) and is not uniform: on MH04 the PRMSE difference is only 0.0005, and on V103 the ROE is worse (0.1931 vs. 0.1884). Please report standard deviations or confidence intervals across independent runs, specify exactly how the 50-sample average is computed (expected error per window versus a single multi-window roll-out), and state the number of training seeds. Without this information, the claim of improved performance is not statistically supported.
  3. [Section V-A, Fig. 2] The claim that predictions are 'more faithful' rests on a single randomly selected one-second window and qualitative inspection. No quantitative bias-level metrics are provided (for example, RMSE or MAE of predicted bias against the reference, correlation, or spectral similarity), and no assessment across multiple windows, sequences, or conditions is given. Given that the reference itself is the VINS proxy used for training, please provide quantitative bias-fidelity metrics on held-out data and, ideally, compare against a more direct reference.
minor comments (5)
  1. [Section IV-A, Eq. (11)] The simplified MSE loss is said to be 'equivalent to the ELBO'; in DDPM this loss is a reweighted variational bound, so the word 'equivalent' is imprecise and should be rephrased.
  2. [Section IV-A, Eq. (13)] Equation (13) uses γ_t without defining it, and the text says DDIM is used for sampling but does not provide the DDIM update rule or how 25 steps are selected from T=1000; please provide the exact schedule and update equations so that the sampling procedure is reproducible.
  3. [Throughout] There are several typographical errors, including 'stochatic' (Abstract), 'uncertainity' (Introduction), 'approxiamte' (Section IV-A), and 'ofinertial-only odometry' (Introduction); these should be corrected.
  4. [Section V-B, Table II] Please clarify whether the reported inference times on the Jetson device include both feature extraction and the 25 DDIM sampling steps, and whether 145 ms is measured for a single one-second input window.
  5. [Section V, Table I] The footnote that V101 was not tested 'as its ground truth accuracy is limited, as reported in [3]' is vague; please specify which ground-truth accuracy issue is meant and why it prevents evaluation.

Circularity Check

1 steps flagged · score 6.0 of 10

The bias-fidelity evaluation is circular: the VINS-recovered, interpolated bias used as training supervision is the same signal used as the Fig. 2 'ground truth,' so matching it only shows fit to the training-target source; the IOO comparison in Table I remains independent.

  1. fitted input called prediction [Section IV-D ('Acquire Bias Ground Truth Data') and Section V-A, Fig. 2]
    "Since IMU bias tends to change slowly over time, the interpolated values offer sufficient accuracy for the use as supervision during training. ... Moreover, we observe that the recovered bias is continuous and changes slowly ... In Fig. 2, our prediction match more closely to the ground truth in both magnitude and changing pattern, In contrast, AirIMU's predictions show abrupt changes, violating our prior knowledge of IMU bias."

    The 'ground truth' bias used to supervise the diffusion model is the same VINS joint-optimization output, interpolated to IMU rate, that later serves as the reference in Fig. 2. The model is trained to reproduce these labels, so matching them in magnitude and variation is a check of fit to the training-target source, not independent evidence that the predictions are physically faithful. The smoothness observation cannot validate the labels because the VINS random-walk prior and interpolation impose slow, continuous changes; the paper's inference that recovered bias 'can serve as effective ground truth' is thus supported only by properties inherited from the estimator.

full rationale

The diffusion training objective itself is standard and not circular: the forward/reverse process and MSE noise loss in Eqs. (10)-(13) are the usual DDPM setup, and the conditional encoder is a normal architectural choice. The main IOO performance comparison in Table I is against external baselines (AirIMU, direct regression, and a random-walk upper-bound baseline) using EuRoC trajectory ground truth, so the central odometry result is not forced by construction. However, the paper's parallel claim of 'more accurate bias prediction' and 'faithful' bias behavior is evaluated against the very signal used as training supervision: the interpolated VINS-recovered bias from Section IV-D. A supervised model is expected to reproduce its training labels on held-out windows if it generalizes at all, so Fig. 2 demonstrates generalization of the learned mapping, not that the labels equal true sensor bias. The assertion that the recovered bias is 'continuous and changes slowly' is also not independent support, because that smoothness is inherited from the VINS random-walk prior and interpolation. This makes the bias-fidelity evaluation partially circular, while leaving the IOO contribution independently supported. I do not count the OpenVINS citation as circular on its own: OpenVINS is an externally evaluated, code-released system, and it is cited alongside OKVIS and VINS-Mono as general sources of bias estimates.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the standard diffusion model objective (from DDPM, cited [30]) and on domain assumptions that VINS-recovered bias is a valid target and that one-second windows provide sufficient context. No new physical entities are introduced.

free parameters (5)
  • Window size = 1 s
    Chosen in Section IV-C to balance context and drift; directly affects all predictions.
  • Overlap ratio = 50%
    Chosen in Section IV-C to balance data diversity and training speed.
  • Diffusion steps T = 1000
    Default from DDPM [30], Section IV-C.
  • DDIM sampling steps = 25
    Chosen in Section IV-C to reduce inference time; affects sample quality.
  • Ensemble size for evaluation = 50
    Section V averages metrics over 50 runs; the mean of samples is used for IOO, a design choice not analyzed.
assumptions (4)
  • standard math The conditional diffusion model trained with MSE noise loss approximates the IMU-conditioned bias distribution.
    Invoked in Section IV-A, following Ho et al. [30].
  • domain assumption Bias recovered from VINS joint optimization is a valid ground truth for IMU bias.
    Section IV-D states the recovered bias is of high quality and supports this via motion integration performance, but it is an estimate, not a direct measurement.
  • domain assumption Bias changes slowly enough that frame-rate estimates can be interpolated to IMU rate.
    Section IV-D; interpolation is used to generate training labels, and any interpolation error is ignored.
  • domain assumption The one-second sliding window with 50% overlap provides sufficient context for bias prediction.
    Section IV-C; the choice follows prior work and is not derived.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Learning IMU Bias with Diffusion Model." pith.science (2026). https://pith.science/paper/EI5BGJL4

@misc{pith2026250511763,
  author       = {Pith},
  title        = {Pith review of: Learning IMU Bias with Diffusion Model},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EI5BGJL4}},
  note         = {Machine review of arXiv:2505.11763}
}
read the original abstract

Motion sensing and tracking with IMU data is essential for spatial intelligence, which however is challenging due to the presence of time-varying stochastic bias. IMU bias is affected by various factors such as temperature and vibration, making it highly complex and difficult to model analytically. Recent data-driven approaches using deep learning have shown promise in predicting bias from IMU readings. However, these methods often treat the task as a regression problem, overlooking the stochatic nature of bias. In contrast, we model bias, conditioned on IMU readings, as a probabilistic distribution and design a conditional diffusion model to approximate this distribution. Through this approach, we achieve improved performance and make predictions that align more closely with the known behavior of bias.

Figures

Figures reproduced from arXiv: 2505.11763 by the authors.

Figure 1
Figure 1. System overview: our model consists of IMU encoder [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Bias prediction result for our model and AirIMU in [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

41 extracted references · 23 canonical work pages

  1. [1]

    Titterton and J

    D. Titterton and J. L. Weston,Strapdown inertial navigation technology. IET, 2004, vol. 17

    2004

  2. [2]

    Unified temporal and spatial calibration for multi-sensor systems,

    P. Furgale, J. Rehder, and R. Siegwart, “Unified temporal and spatial calibration for multi-sensor systems,” in2013 IEEE/RSJ International Conference on Intelligent Robots and Systems. IEEE, 2013, pp. 1280– 1286

    work page 2013

  3. [3]

    OpenVINS: a research platform for visual-inertial estimation,

    P. Geneva, K. Eckenhoff, W. Lee, Y . Yang, and G. Huang, “OpenVINS: a research platform for visual-inertial estimation,” inProc. of the IEEE International Conference on Robotics and Automation, Paris, France, 2020. [Online]. Available: https://github.com/rpng/open vins

    work page 2020

  4. [4]

    Inertial navigation meets deep learning: A survey of current trends and future directions,

    N. Cohen and I. Klein, “Inertial navigation meets deep learning: A survey of current trends and future directions,”Results in Engineering, p. 103565, 2024

    2024

  5. [5]

    Imu data processing for inertial aided navigation: A recurrent neural network based approach,

    M. Zhang, M. Zhang, Y . Chen, and M. Li, “Imu data processing for inertial aided navigation: A recurrent neural network based approach,” in2021 IEEE International Conference on Robotics and Automation (ICRA), 2021, pp. 3992–3998

    work page 2021

  6. [6]

    Airimu: Learning uncertainty propagation for inertial odometry,

    Y . Qiu, C. Wang, C. Xu, Y . Chen, X. Zhou, Y . Xia, and S. Scherer, “Airimu: Learning uncertainty propagation for inertial odometry,”arXiv preprint arXiv:2310.04874, 2023

    arXiv 2023

  7. [7]

    Deep IMU Bias Inference for Robust Visual-Inertial Odometry With Factor Graphs,

    R. Buchanan, V . Agrawal, M. Camurri, F. Dellaert, and M. Fallon, “Deep IMU Bias Inference for Robust Visual-Inertial Odometry With Factor Graphs,”IEEE Robotics and Automation Letters, vol. 8, no. 1, pp. 41–48, Jan. 2023

    work page 2023

  8. [8]

    Diffusion policy: Visuomotor policy learning via action diffusion,

    C. Chi, Z. Xu, S. Feng, E. Cousineau, Y . Du, B. Burchfiel, R. Tedrake, and S. Song, “Diffusion policy: Visuomotor policy learning via action diffusion,”The International Journal of Robotics Research, 2024

    2024

Show all 41 references
  1. [9]

    Stochastic trajectory prediction via motion indeterminacy diffusion,

    T. Gu, G. Chen, J. Li, C. Lin, Y . Rao, J. Zhou, and J. Lu, “Stochastic trajectory prediction via motion indeterminacy diffusion,” inProceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 17 113–17 122

    2022

  2. [10]

    Walk detection and step counting on unconstrained smartphones,

    R. H. A. Brajdi, “Walk detection and step counting on unconstrained smartphones,” inACM international joint conference on Pervasive and ubiquitous computing, 2011

    2011

  3. [11]

    Pedestrian tracking with shoe-mounted inertial sensors,

    E. Foxlin, “Pedestrian tracking with shoe-mounted inertial sensors,” IEEE Computer graphics and applications, vol. 25, no. 6, pp. 38–46, 2005

    2005

  4. [12]

    Inertial Odometry on Handheld Smartphones,

    A. Solin, S. Cortes, E. Rahtu, and J. Kannala, “Inertial Odometry on Handheld Smartphones,” in2018 21st International Conference on Information Fusion (FUSION). Cambridge, United Kingdom: IEEE, July 2018, pp. 1–5

    2018

  5. [13]

    Ridi: Robust imu double integration,

    H. Yan, Q. Shan, and Y . Furukawa, “Ridi: Robust imu double integration,” inProceedings of the European conference on computer vision (ECCV), 2018, pp. 621–636

    2018

  6. [14]

    Ionet: Learning to cure the curse of drift in inertial odometry,

    C. Chen, X. Lu, A. Markham, and N. Trigoni, “Ionet: Learning to cure the curse of drift in inertial odometry,” inProceedings of the AAAI Conference on Artificial Intelligence, vol. 32, no. 1, 2018

    2018

  7. [15]

    Idol: Inertial deep orientation- estimation and localization,

    S. Sun, D. Melamed, and K. Kitani, “Idol: Inertial deep orientation- estimation and localization,” inProceedings of the AAAI Conference on Artificial Intelligence, vol. 35, no. 7, 2021, pp. 6128–6137

    2021

  8. [16]

    Ronin: Robust neural inertial navigation in the wild: Benchmark, evaluations, & new methods,

    S. Herath, H. Yan, and Y . Furukawa, “Ronin: Robust neural inertial navigation in the wild: Benchmark, evaluations, & new methods,” in 2020 IEEE international conference on robotics and automation (ICRA). IEEE, 2020, pp. 3146–3152

    2020

  9. [17]

    Tlio: Tight learned inertial odometry,

    W. Liu, D. Caruso, E. Ilg, J. Dong, A. I. Mourikis, K. Daniilidis, V . R. Kumar, and J. J. Engel, “Tlio: Tight learned inertial odometry,”IEEE Robotics and Automation Letters, vol. 5, pp. 5653–5660, 2020

    2020

  10. [18]

    Neural inertial localization,

    S. Herath, D. Caruso, C. Liu, Y . Chen, and Y . Furukawa, “Neural inertial localization,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 6604–6613

    2022

  11. [19]

    Learned inertial odometry for autonomous drone racing,

    G. Cioffi, L. Bauersfeld, E. Kaufmann, and D. Scaramuzza, “Learned inertial odometry for autonomous drone racing,”IEEE Robotics and Automation Letters, vol. 8, no. 5, pp. 2684–2691, 2023

    2023

  12. [20]

    Rio: Rotation-equivariance supervised learning of robust inertial odometry,

    X. Cao, C. Zhou, D. Zeng, and Y . Wang, “Rio: Rotation-equivariance supervised learning of robust inertial odometry,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 6614–6623

    2022

  13. [21]

    Eqnio: Subequivariant neural inertial odometry,

    R. K. Jayanth, Y . Xu, Z. Wang, E. Chatzipantazis, D. Gehrig, and K. Daniilidis, “Eqnio: Subequivariant neural inertial odometry,”arXiv preprint arXiv:2408.06321, 2024

    2024 arXiv

  14. [22]

    Indirect kalman filter for 3d attitude estimation,

    N. Trawny and S. I. Roumeliotis, “Indirect kalman filter for 3d attitude estimation,”University of Minnesota, Dept. of Comp. Sci. & Eng., Tech. Rep, vol. 2, p. 2005, 2005

    2005

  15. [23]

    Denoising imu gyroscopes with deep learning for open-loop attitude estimation,

    S. Brossard, Martin ands Bonnabel and A. Barrau, “Denoising imu gyroscopes with deep learning for open-loop attitude estimation,”IEEE Robotics and Automation Letters, vol. 5, no. 3, pp. 4796–4803, 2020

    2020

  16. [24]

    Visual-inertial navigation: A concise review,

    G. Huang, “Visual-inertial navigation: A concise review,” inProc. International Conference on Robotics and Automation, Montreal, Canada, May 2019

    2019

  17. [25]

    Observability analysis of aided ins with heterogeneous features of points, lines and planes,

    Y . Yang and G. Huang, “Observability analysis of aided ins with heterogeneous features of points, lines and planes,”IEEE Transactions on Robotics, vol. 35, no. 6, pp. 399–1418, Dec. 2019

    2019

  18. [26]

    Low-frequency spectral analysis of dc nanovoltmeters and voltage reference standards,

    T. Witt, “Low-frequency spectral analysis of dc nanovoltmeters and voltage reference standards,”IEEE Transactions on Instrumentation and Measurement, vol. 46, no. 2, pp. 318–321, 1997

    1997

  19. [27]

    Should the classical variance be used as a basic measure in standards metrology?

    D. W. Allan, “Should the classical variance be used as a basic measure in standards metrology?”IEEE Transactions on Instrumentation and Measurement, vol. IM-36, no. 2, pp. 646–654, 1987

    1987

  20. [28]

    Ieee standard specification format guide and test procedure for coriolis vibratory gyros,

    “Ieee standard specification format guide and test procedure for coriolis vibratory gyros,”IEEE Std 1431-2004, pp. 1–78, 2004

    2004

  21. [29]

    Ieee standard specification format guide and test procedure for linear single-axis, nongyroscopic accelerometers,

    “Ieee standard specification format guide and test procedure for linear single-axis, nongyroscopic accelerometers,”IEEE Std 1293-2018 (Revision of IEEE Std 1293-1998), pp. 1–271, 2019

    2018

  22. [30]

    Denoising diffusion probabilistic models,

    J. Ho, A. Jain, and P. Abbeel, “Denoising diffusion probabilistic models,” Advances in neural information processing systems, vol. 33, pp. 6840– 6851, 2020

    2020

  23. [31]

    An empirical evaluation of generic convolutional and recurrent networks for sequence modeling,

    S. Bai, J. Z. Kolter, and V . Koltun, “An empirical evaluation of generic convolutional and recurrent networks for sequence modeling,” arXiv:1803.01271, 2018

    2018 arXiv

  24. [32]

    Attention is all you need,

    A. Vaswani, “Attention is all you need,”Advances in Neural Information Processing Systems, 2017

    2017

  25. [33]

    U-net: Convolutional net- works for biomedical image segmentation,

    O. Ronneberger, P. Fischer, and T. Brox, “U-net: Convolutional net- works for biomedical image segmentation,” inMedical image computing and computer-assisted intervention–MICCAI 2015: 18th international conference, Munich, Germany, October 5-9, 2015, proceedings, part III 18. ...

    2015

  26. [34]

    Learning phrase representations using RNN encoder–decoder for statistical machine translation,

    K. Cho, B. van Merri ¨enboer, C. Gulcehre, D. Bahdanau, F. Bougares, H. Schwenk, and Y . Bengio, “Learning phrase representations using RNN encoder–decoder for statistical machine translation,” inProceed- ings of the 2014 Conference on Empirical Methods in Natural Language Pro...

    2014

  27. [35]

    Adam: A method for stochastic optimization,

    D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” in3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings, Y . Bengio and Y . LeCun, Eds., 2015

    2015

  28. [36]

    Denoising diffusion implicit models,

    J. Song, C. Meng, and S. Ermon, “Denoising diffusion implicit models,” arXiv preprint arXiv:2010.02502, 2020

    2010 arXiv

  29. [37]

    Vision and imu data fusion: Closed-form solutions for attitude, speed, absolute scale, and bias determination,

    A. Martinelli, “Vision and imu data fusion: Closed-form solutions for attitude, speed, absolute scale, and bias determination,”IEEE Transactions on Robotics, vol. 28, no. 1, pp. 44–60, 2012

    2012

  30. [38]

    Keyframe-based visual–inertial odometry using nonlinear optimization,

    S. Leutenegger, S. Lynen, M. Bosse, R. Siegwart, and P. Furgale, “Keyframe-based visual–inertial odometry using nonlinear optimization,” The International Journal of Robotics Research, vol. 34, no. 3, pp. 314–334, 2015

    2015

  31. [39]

    Vins-mono: A robust and versatile monoc- ular visual-inertial state estimator,

    T. Qin, P. Li, and S. Shen, “Vins-mono: A robust and versatile monoc- ular visual-inertial state estimator,”IEEE transactions on robotics, vol. 34, no. 4, pp. 1004–1020, 2018

    2018

  32. [40]

    The newer college dataset: Handheld lidar, inertial and vision with ground truth,

    M. Ramezani, Y . Wang, M. Camurri, D. Wisth, M. Mattamala, and M. Fallon, “The newer college dataset: Handheld lidar, inertial and vision with ground truth,” in2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020, pp. 4353–4360

    2020

  33. [41]

    The euroc micro aerial vehicle datasets,

    M. Burri, J. Nikolic, P. Gohl, T. Schneider, J. Rehder, S. Omari, M. W. Achtelik, and R. Siegwart, “The euroc micro aerial vehicle datasets,” The International Journal of Robotics Research, 2016

    2016

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.