Pith. sign in

REVIEW 3 major objections 4 minor 1 cited by

Learning to Double Guess: An Active Perception Approach for Estimating the Center of Mass of Arbitrary Objects

T0 review · 3 major / 4 minor · reviewed 2026-08-09 · deepseek-v4-flash

Pith's one-line read A robot can estimate an unseen object's center of mass by taking two force-torque measurements, with the second orientation picked by a learned uncertainty-guided scorer, reaching 1.47 cm average error after training on just two objects.

desk verdict Useful active-perception CoM paper with real experiments, but the fusion step in Sec III-D has a frame-consistency ambiguity that needs fixing, and the abstract overclaims relative to the limitations. read the letter →

arxiv 2502.02663 v2 pith:LDRORW2H submitted 2025-02-04 cs.RO cs.LG

classification cs.ROcs.LG
keywords centerofmassestimationactiveperceptionBayesianneuralnetworkforce-torquesensinguncertaintyquantificationrobotmanipulationzero-shotgeneralizationhaptic
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

This paper claims that a robot can estimate the center of mass of an arbitrary rigid object it has never seen before by taking two force-torque measurements at two different grasp orientations. The first measurement feeds a Bayesian neural network that outputs a mean and a standard deviation for the center-of-mass offset; a second network, ActiveNet, scores every candidate rotation angle by how much it would reduce the expected error, and the robot rotates to the best one and measures again. The two estimates are fused as an inverse-variance-weighted average, which the authors show is a simple Gaussian posterior. They report an average error of 1.47 cm (7.6%) on twelve everyday objects, after training on only two custom 3D-printed objects loaded with laboratory weights. If this transfer is reliable, it would give robots a general way to perceive a physical property that is usually assumed known for grasping and manipulation.

What carries the argument

The framework's two load-bearing components are a Bayesian Neural Network (BNN) trained with Markov Chain Monte Carlo using the No-U-Turn Sampler, and ActiveNet, a 5-layer MLP that scores candidate second orientations. The BNN gives each prediction a mean and standard deviation, which is the uncertainty that drives action selection. ActiveNet takes the first estimate's mean and standard deviation plus a candidate second rotation, and predicts the error of the next measurement; a grid search over the continuous 2D orientation space chooses the minimum-score rotation. The final estimate is the inverse-variance-weighted mean of the two Gaussian estimates from the two orientations, treating them as independent measurements.

What would settle it

Collect a test set of objects spanning weights from 50 g to 800 g, including low-friction surfaces and shapes unlike the training plate and box, and run both U-GRAPH and the Random Rotate baseline on each. If U-GRAPH does not beat Random Rotate on the z-axis for a majority of objects, or if its average error on objects outside the 127–585 g training range exceeds 3 cm, the zero-shot generalization and uncertainty-guided action-selection claims would be contradicted.

Watch

Extended reading notes

Core claim

The core discovery is that a mapping from 6-axis force-torque readings and grasp orientation to the 3D center-of-mass offset can be learned on a tiny dataset with limited variation and still transfer to arbitrary rigid objects in zero-shot fashion. The model outputs a Gaussian predictive distribution, so the robot knows not only where the center of mass is but how sure it is. That uncertainty is then used to choose a second rotation angle that promises the largest information gain, and the two Gaussian estimates are combined into a posterior. Against one-measurement baselines, the second measurement is what makes z-axis estimation possible, and against a random second rotation, the uncertainty-guided choice consistently yields lower error in the paper's experiments.

Load-bearing premise

The mapping between force-torque readings and center-of-mass offsets is learned only from two custom plastic objects with laboratory weights in a narrow range, and the generalization claim assumes that this mapping transfers to arbitrary real-world objects despite differences in geometry, friction, density, and weight.

Editorial extensions

If this is right

  • A single vertical grasp cannot determine the center-of-mass offset along the gravity axis; taking a second, uncertainty-guided rotation always improves the z-axis estimate in the paper's experiments.
  • Objects whose weights lie inside the training range (127–585 g) are estimated with a few millimeters of error; objects outside that range show noticeably larger errors, so the learned mapping is weight-sensitive.
  • The trained model transfers to objects with different contact geometry, surface friction, and density, suggesting that the force-torque-to-CoM mapping captures something general about statics rather than memorizing the training shapes.
  • The two-measurement approach replaces repeated re-grasping with an in-hand rotation, which is faster and safer for delicate objects.
  • Extending the pipeline to more than two actions, with recursive posterior updates, is the paper's own stated next step and would likely improve accuracy further.

Reading between the lines

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

  • The success of uncertainty-guided action selection suggests that an ensemble of deterministic networks, which also provides predictive variance, might achieve similar gains at lower training cost; the paper's BNN provides a natural baseline for that comparison.
  • The inverse-variance fusion treats the two measurements as independent Gaussians, but the second measurement is physically correlated with the first through the shared object; modelling that correlation could improve the posterior and further reduce error.
  • Because the reported 1.47 cm is an average, its practical value depends on the error distribution; the paper's own data show larger errors on out-of-distribution weights, so deployment should include a weight check or an adapted training set.
  • The same active-perception loop could be applied to other physical parameters that are observable only through contact, such as friction or compliance, whenever a model can output uncertainty.
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 / 4 minor

Summary. This paper presents U-GRAPH, a two-measurement active perception system for estimating the 3D center of mass of grasped objects. A Bayesian neural network (BNN) maps a 6D force-torque reading and a 2D gripper orientation to a CoM offset and an uncertainty estimate; an ActiveNet, queried by grid search, scores candidate second orientations using the BNN's first-prediction mean and variance, and the system rotates to the best-scoring orientation, takes a second F/T reading, and fuses the two BNN predictions via inverse-variance weighting. The BNN is trained only on two 3D-printed objects with attached laboratory weights, yet the authors report an average error of 1.47 cm (7.6%) on 12 unseen real-world objects and claim zero-shot generalization, with U-GRAPH outperforming one-grasp, analytical, and random-rotation baselines on most objects.

Significance. If validated, this work addresses a genuine and under-explored problem: estimating the 3D CoM of arbitrary rigid objects from haptic interaction, which is useful for stable grasping and manipulation. The use of a Bayesian neural network to produce predictive uncertainty and the formulation of the second rotation as an information-gathering action are reasonable and testable. The evaluation on 12 real objects with independently measured CoM locations, including out-of-distribution weights, is a strength, and the reported OOD degradation in Table II and Sec. V-E is honest. The posterior-combination step is not circular, since the BNN is trained with ground-truth labels and evaluated on unseen objects. However, the frame convention for the fused estimates and the lack of variance reporting prevent me from accepting the quantitative claims as stated.

major comments (3)
  1. [Sec. III-A, Sec. III-D (Eq. 1)] The fusion in Eq. (1) assumes mu_1 and mu_2 are estimates of the same vector. Section III-A defines dx, dy, dz as world-frame displacements from the grasping point, and the paper does not state that the BNN output for a rotated pose is transformed back to the initial grasp frame before fusion. If the outputs are world-frame, after a gripper rotation R(theta) the physical CoM offset in the world frame is R(theta)*r, not r, and averaging mu_1 and mu_2 without rotating mu_2 back introduces a bias of order (R(theta)-I)*r/2, which can be several centimeters for the large angles used to make the z-offset observable. If the intended convention is that the BNN always outputs the CoM in the initial grasp frame (which the data-collection procedure in Sec. IV-B makes plausible, since labels are computed before rotation), this must be stated explicitly and the training labels must be confirmed to use that convention. As written, the ambiguity is load-bearing for the headline 1.47 cm result and for the comparison with Random Rotate; please state the convention, or apply an explicit rotation correction, and report the results under both conventions.
  2. [Sec. V-C, Table I] Table I reports only per-axis mean errors over five grasps, with no variance or per-grasp breakdown, and the aggregate 1.47 cm / 7.6% accuracy quoted in the abstract is not reproducible from the table. Without error bars or standard errors, the claim that U-GRAPH outperforms Random Rotate cannot be assessed statistically; several entries differ by only a few millimeters. Please report per-grasp 3D errors, standard deviations or confidence intervals, and define the denominator used for the 7.6% figure.
  3. [Sec. V-A] The analytical baseline formula is written as r = tau × F / |F|^2. The correct inversion of tau = r × F is r = (F × tau) / |F|^2; as written, the formula has the cross-product order reversed. If the implemented baseline used the printed formula, its errors are not meaningful; if it used the correct formula, the text should be corrected. Either way, the baseline comparison needs to be based on the correct formula.
minor comments (4)
  1. [Sec. I] In the Introduction, 'This work provides a framework for perceiving the CoM of an arbitrary object. object.' contains a duplicated word; please fix the typo.
  2. [Sec. III-A] The citation 'as illustrated in Fig. III-A' should refer to the actual figure number (Fig. 2).
  3. [Sec. III-D] Equation (1) is typeset awkwardly (the fraction is inline and difficult to read); consider displaying it with standard fraction formatting.
  4. [Sec. V-E] The claim of generalization to 'arbitrary rigid object[s]' in the abstract is stronger than the evidence: Table II and the OOD rows of Table I show clear degradation outside the training weight range (43.4 g, 76.8 g, 613.2 g). Please qualify the phrasing, e.g., 'objects within the approximate weight range of the training data'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the CoM estimates and the active-rotation improvement are evaluated against external ground truth on held-out objects.

full rationale

The paper's central chain is empirical rather than definitional. The BNN is trained on F/T readings and orientation inputs with CoM labels computed from AprilTag detections on two custom objects; the test claims are evaluated on 12 unseen real-world objects whose ground-truth CoM is obtained by independent balancing, so the reported errors are not re-derivations of training labels. The ActiveNet is trained to predict the BNN's own error at candidate second orientations, but at inference it only selects an orientation; the final CoM estimate comes from the BNN posterior fusion in Eq. (1) and does not use the ActiveNet score or any ground-truth information. The comparison against Random Rotate is therefore a meaningful empirical test. The posterior Gaussian fusion is a standard combination of two independently noisy estimates. The paper's self-citations (e.g., [2], [3], [4], [6], [11], [22], [30]) appear only in related-work or future-work contexts and are not load-bearing for the central derivation. The potential world-frame versus gripper-frame inconsistency in Eq. (1) that the skeptic note raises is a correctness concern about coordinate transforms, not a circularity: it does not make the output equal to an input by construction. Hence no circular step is present.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new physical entities. The central claim rests on learned network weights, a domain assumption about the F/T-to-CoM mapping, and the transferability of that mapping to unseen objects. The free parameters are the neural network weights and a few hand-chosen hyperparameters.

free parameters (3)
  • BNN weights = not disclosed
    Trained on 204 grasps from two objects; the learned mapping from F/T readings to CoM offsets is the core fitted model.
  • ActiveNet weights = not disclosed
    Trained to predict the BNN's estimation error for candidate actions; this network guides the active perception.
  • BNN initialization std = 0.5
    Hand-picked in Sec IV-C as the standard deviation for initializing the BNN from a pretrained deterministic MLP.
assumptions (4)
  • domain assumption The CoM of any grasped object is determined by a fixed offset (dx, dy, dz) and gravitational force, and the F/T reading at a given orientation is a noisy function of these.
    This is the intuitive model in Sec III-A and the basis for defining the regression targets.
  • domain assumption Two measurements from different orientations are independent and Gaussian, so the posterior combination formula in Sec III-D is valid.
    The final estimate uses the closed-form Gaussian fusion formula without validating the independence or Gaussianity assumptions.
  • domain assumption The BNN trained on two custom objects transfers to arbitrary rigid objects despite differences in geometry, friction, and density.
    The generalization claim in Sec V-C rests on this transferability, which is tested empirically but not explained mechanistically.
  • domain assumption The AprilTag-based ground truth (training) and the balancing method (test) provide accurate CoM measurements.
    The supervised labels and evaluation targets both depend on the accuracy of these ground-truth procedures, described in Sec IV-B and Sec V-C.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Learning to Double Guess: An Active Perception Approach for Estimating the Center of Mass of Arbitrary Objects." pith.science (2026). https://pith.science/paper/LDRORW2H

@misc{pith2026250202663,
  author       = {Pith},
  title        = {Pith review of: Learning to Double Guess: An Active Perception Approach for Estimating the Center of Mass of Arbitrary Objects},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LDRORW2H}},
  note         = {Machine review of arXiv:2502.02663}
}
read the original abstract

Manipulating arbitrary objects in unstructured environments is a significant challenge in robotics, primarily due to difficulties in determining an object's center of mass. This paper introduces U-GRAPH: Uncertainty-Guided Rotational Active Perception with Haptics, a novel framework to enhance the center of mass estimation using active perception. Traditional methods often rely on single interaction and are limited by the inherent inaccuracies of Force-Torque (F/T) sensors. Our approach circumvents these limitations by integrating a Bayesian Neural Network (BNN) to quantify uncertainty and guide the robotic system through multiple, information-rich interactions via grid search and a neural network that scores each action. We demonstrate the remarkable generalizability and transferability of our method with training on a small dataset with limited variation yet still perform well on unseen complex real-world objects.

Figures

Figures reproduced from arXiv: 2502.02663 by the authors.

Figure 1
Figure 1. We design an active perception algorithm to estimate [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Illustration of the simplified model of CoM on a [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. a) Flowchart for training Bayesian Neural Network. We train BNN with Markov Chain Monte Carlo and No U-Turn [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: a) Example of a data collection robot grasping with [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 5
Figure 5. Figure 5: Mean error and mean standard deviation (shown with [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Rapid Mismatch Estimation via Neural Network Informed Variational Inference

    cs.RO 2025-08 conditional novelty 6.0 of 10

    RME estimates end-effector mass and center-of-mass mismatches online in about 400 ms using proprioceptive feedback and a neural-network-guided variational inference.

Reference graph

Works this paper leans on

30 extracted references · 27 canonical work pages · cited by 1 Pith paper

  1. [1]

    Kosmos-e : Learning to follow instruction for robotic grasping,

    Z. Wang, X. Wu, S. Huang, L. Dong, W. Wang, S. Ma, and F. Wei, “Kosmos-e : Learning to follow instruction for robotic grasping,” in 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2024, pp. 9510–9517

  2. [2]

    Towards open-world interactive disambiguation for robotic grasping,

    Y . Mo, H. Zhang, and T. Kong, “Towards open-world interactive disambiguation for robotic grasping,” in 2023 IEEE International Conference on Robotics and Automation (ICRA) , 2023, pp. 8061– 8067

  3. [3]

    Poseit: A visual-tactile dataset of holding poses for grasp stability analysis,

    S. Kanitkar, H. Jiang, and W. Yuan, “Poseit: A visual-tactile dataset of holding poses for grasp stability analysis,” in 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2022, pp. 71–78

  4. [4]

    Grasp stability prediction with sim-to-real transfer from tactile sensing,

    Z. Si, Z. Zhu, A. Agarwal, S. Anderson, and W. Yuan, “Grasp stability prediction with sim-to-real transfer from tactile sensing,” in 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2022, pp. 7809–7816

  5. [5]

    A Practical Bayesian Framework for Backpropa- gation Networks,

    D. J. C. MacKay, “A Practical Bayesian Framework for Backpropa- gation Networks,” Neural Computation, vol. 4, no. 3, pp. 448–472, 05 1992

  6. [6]

    Understanding Dynamic Tactile Sensing for Liquid Property Estimation,

    H.-J. Huang, X. Guo, and W. Yuan, “Understanding Dynamic Tactile Sensing for Liquid Property Estimation,” in Proceedings of Robotics: Science and Systems , New York City, NY , USA, June 2022

  7. [7]

    Estimating properties of solid particles inside container using touch sensing,

    X. Guo, H.-J. Huang, and W. Yuan, “Estimating properties of solid particles inside container using touch sensing,” in 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2023, pp. 8985–8992

  8. [8]

    Haptic perception of liquids enclosed in containers,

    C. Matl, R. Matthew, and R. Bajcsy, “Haptic perception of liquids enclosed in containers,” in 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2019, pp. 7142–7149

Show all 30 references
  1. [9]

    Stir to pour: Efficient calibration of liquid properties for pouring actions,

    T. Lopez-Guevara, R. Pucci, N. K. Taylor, M. U. Gutmann, S. Ra- mamoorthy, and K. Suhr, “Stir to pour: Efficient calibration of liquid properties for pouring actions,” in 2020 IEEE/RSJ International Con- ference on Intelligent Robots and Systems (IROS) , 2020, pp. 5351– 5357

  2. [10]

    Kitchen artist: Precise control of liquid dispensing for gourmet plating,

    H.-J. Huang, J. Xiang, and W. Yuan, “Kitchen artist: Precise control of liquid dispensing for gourmet plating,” in 2024 IEEE International Conference on Robotics and Automation (ICRA) , 2024, pp. 13 933– 13 939

  3. [11]

    An intelligent robotic system for perceptive pancake batter stirring and precise pouring,

    X. Luo, S. Jin, H.-J. Huang, and W. Yuan, “An intelligent robotic system for perceptive pancake batter stirring and precise pouring,” in 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2024, pp. 5970–5977

  4. [12]

    Toss- ingbot: Learning to throw arbitrary objects with residual physics,

    A. Zeng, S. Song, J. Lee, A. Rodriguez, and T. Funkhouser, “Toss- ingbot: Learning to throw arbitrary objects with residual physics,” in Proceedings of Robotics: Science and Systems (RSS) , 2019

  5. [13]

    Swing- bot: Learning physical features from in-hand tactile exploration for dynamic swing-up manipulation,

    C. Wang, S. Wang, B. Romero, F. Veiga, and E. Adelson, “Swing- bot: Learning physical features from in-hand tactile exploration for dynamic swing-up manipulation,” in 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020, pp. 5633– 5640

  6. [14]

    Simultaneous planning and estimation based on physics reasoning in robot manipulation,

    M. Murooka, S. Nozawa, M. Bando, I. Yanokura, K. Okada, and M. Inaba, “Simultaneous planning and estimation based on physics reasoning in robot manipulation,” in 2018 IEEE International Confer- ence on Robotics and Automation (ICRA) , 2018, pp. 3137–3144

  7. [15]

    Predicting center of mass by iterative pushing for object transportation and manipulation,

    S. M. Hyland, J. Xiao, and C. D. Onal, “Predicting center of mass by iterative pushing for object transportation and manipulation,” in 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2023, pp. 1615–1620

  8. [16]

    Learning to estimate centers of mass of arbitrary objects,

    S. McGovern, H. Mao, and J. Xiao, “Learning to estimate centers of mass of arbitrary objects,” in 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2019, pp. 1848–1853

  9. [17]

    Learning and predicting center of mass through manipulation and torque sensing,

    S. McGovern and J. Xiao, “Learning and predicting center of mass through manipulation and torque sensing,” in 2022 8th International Conference on Mechatronics and Robotics Engineering (ICMRE) , 2022, pp. 60–66

  10. [18]

    Observations on active touch

    J. J. Gibson, “Observations on active touch.” Psychological Review, vol. 69, pp. 477–491, 1962

  11. [19]

    Tandem3d: Active tactile exploration for 3d object recognition,

    J. Xu, H. Lin, S. Song, and M. Ciocarlie, “Tandem3d: Active tactile exploration for 3d object recognition,” in 2023 IEEE International Conference on Robotics and Automation (ICRA) , 2023, pp. 10 401– 10 407

  12. [20]

    Active haptic exploration based on dual-stage perception for object recognition,

    P. Uttayopas, X. Cheng, and E. Burdet, “Active haptic exploration based on dual-stage perception for object recognition,” in 2023 IEEE World Haptics Conference (WHC) , 2023, pp. 347–353

  13. [21]

    Interactive learning of physical object properties through robot manipulation and database of object measurements,

    A. Kruzliak, J. Hartvich, S. P. Patni, L. Rustler, J. K. Behrens, F. J. Abu-Dakka, K. Mikolajczyk, V . Kyrki, and M. Hoffmann, “Interactive learning of physical object properties through robot manipulation and database of object measurements,” in 2024 IEEE/RSJ International Co...

  14. [22]

    Active clothing material perception using tactile sensing and deep learning,

    W. Yuan, Y . Mo, S. Wang, and E. H. Adelson, “Active clothing material perception using tactile sensing and deep learning,” in 2018 IEEE International Conference on Robotics and Automation (ICRA) , 2018, pp. 4842–4849

  15. [23]

    Active exploration for real-time haptic training,

    J. Ketchum, A. Prabhakar, and T. D. Murphey, “Active exploration for real-time haptic training,” in 2024 IEEE International Conference on Robotics and Automation (ICRA) , 2024, pp. 9689–9695

  16. [24]

    Weight uncertainty in neural networks,

    C. Blundell, J. Cornebise, K. Kavukcuoglu, and D. Wierstra, “Weight uncertainty in neural networks,” in Proceedings of the 32nd Interna- tional Conference on International Conference on Machine Learning - Volume 37, ser. ICML’15, 2015, p. 1613–1622

  17. [25]

    The no-u-turn sampler: Adaptively setting path lengths in hamiltonian monte carlo,

    M. D. Hoffman and A. Gelman, “The no-u-turn sampler: Adaptively setting path lengths in hamiltonian monte carlo,” 2011. [Online]. Available: https://arxiv.org/abs/1111.4246

  18. [26]

    Brooks, A

    S. Brooks, A. Gelman, G. Jones, and X.-L. Meng, Handbook of Markov Chain Monte Carlo . Chapman and Hall/CRC, May 2011. [Online]. Available: http://dx.doi.org/10.1201/b10905

  19. [27]

    Pyro: deep universal probabilistic programming,

    E. Bingham, J. P. Chen, M. Jankowiak, F. Obermeyer, N. Pradhan, T. Karaletsos, R. Singh, P. Szerlip, P. Horsfall, and N. D. Goodman, “Pyro: deep universal probabilistic programming,” J. Mach. Learn. Res., vol. 20, no. 1, p. 973–978, jan 2019

  20. [28]

    AprilTag: A robust and flexible visual fiducial system,

    E. Olson, “AprilTag: A robust and flexible visual fiducial system,” in Proceedings of the IEEE International Conference on Robotics and Automation (ICRA). IEEE, May 2011, pp. 3400–3407

  21. [29]

    On the variance of the adaptive learning rate and beyond,

    L. Liu, H. Jiang, P. He, W. Chen, X. Liu, J. Gao, and J. Han, “On the variance of the adaptive learning rate and beyond,” in Proceedings of the Eighth International Conference on Learning Representations (ICLR 2020), April 2020

  22. [30]

    Gelsight: High- resolution robot tactile sensors for estimating geometry and force,

    W. Yuan, S. Dong, and E. H. Adelson, “Gelsight: High- resolution robot tactile sensors for estimating geometry and force,” Sensors, vol. 17, no. 12, 2017. [Online]. Available: https://www.mdpi.com/1424-8220/17/12/2762

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.