Pith. sign in

REVIEW 4 major objections 7 minor 20 references

DOGE: An Extrinsic Orientation and Gyroscope Bias Estimation for Visual-Inertial Odometry Initialization

T0 review · 4 major / 7 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read A rotation-only constraint can estimate the camera-IMU extrinsic rotation and gyroscope bias together during visual-inertial initialization, before translation parallax exists.

desk verdict A genuinely new rotation-only extrinsic/gyro-bias estimator with strong experiments, but the pure-rotation claim lacks an observability condition and overreaches. read the letter →

arxiv 2412.08135 v1 pith:TSXO2JK3 submitted 2024-12-11 cs.RO cs.CVcs.LG

Zewen Xu , Yijia He , Hao Wei , Yihong Wu This is my paper · ORCID
classification cs.ROcs.CVcs.LG
keywords visual-inertialinitializationextrinsicorientationestimationgyroscopebiasnormalepipolarconstraintsrotation-onlymaximumaposterioriiterativelyreweightedleastsquaresfailuredetection
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 claims that the camera-IMU extrinsic rotation and the gyroscope bias can be estimated together during visual-inertial initialization using only rotational motion, before translation parallax exists. The method, called DOGE, builds this estimate from a normal epipolar constraint that depends only on rotation, coupling gyroscope preintegration with the visual bearing vectors. The authors argue that this removes the need to wait for translation excitation and avoids the risk of bundle-adjustment-based post-refinement failing when the initial extrinsic guess is poor. In experiments, when the initial extrinsic rotation error exceeds 5 degrees, DOGE produces lower rotation and bias errors than the compared baselines, and the full VIO system starts with a better trajectory estimate.

What carries the argument

The load-bearing object is the normal epipolar constraint (NEC), which replaces the standard epipolar constraint with the coplanarity of epipolar-plane normal vectors, making the two-view cost depend only on rotation through the minimum eigenvalue of the matrix of normal-vector outer products. DOGE writes the rotation between two camera frames as the extrinsic rotation conjugated with the gyroscope preintegration, then linearizes that expression in the gyroscope bias error and the extrinsic orientation error. That linearized rotation is inserted into the NEC cost, and the minimization of the sum of minimum eigenvalues over keyframe pairs is solved with iteratively reweighted least squares, using an eigenvalue-based weight, a per-feature-pair weight derived from propagated pixel and IMU covariance, and a chi-square test to reject outliers. A maximum a posteriori step, implemented with an iterated error-state Kalman filter, carries the estimates across sliding windows before translation parallax arrives.

What would settle it

Run the estimator on a synthetic sequence with pure rotation and zero translation; if it returns biased values even with zero pixel noise, or if the cost is flat along some direction, the rotation-only constraint is not identifiable. A cheaper check is to evaluate the observability matrix of the linearized constraint at a pure-rotation trajectory and inspect its null space.

Watch

Extended reading notes

Core claim

The central claim is that the six-dimensional error state consisting of gyroscope bias correction and extrinsic orientation error is identifiable from normal epipolar constraints alone, even in pure-rotation motion. The paper rewrites the inter-frame rotation as a function of the gyroscope measurements and the extrinsic rotation, linearizes the product in the two error terms, and substitutes this expression into the NEC minimum-eigenvalue cost. The resulting cost is minimized by iteratively reweighted least squares with a chi-square outlier test, and a maximum a posteriori estimator then propagates and refines the estimates across sliding windows until translation parallax accumulates. The paper shows that this yields accurate extrinsic rotation and gyroscope bias estimates with fewer keyframes than methods that wait for translation or that post-optimize with visual-inertial bundle adjustment.

Load-bearing premise

The method assumes that the cost built from two keyframe pairs and a first-order linearization pins down all six unknowns (gyroscope bias and extrinsic rotation) even when the camera is not moving sideways; if that cost is flat or ambiguous, the estimator cannot converge.

Editorial extensions

If this is right

  • If the rotation-only constraint is identifiable, VIO initialization no longer needs to wait for translation excitation, cutting startup latency in applications such as AR and drone navigation.
  • Joint estimation of gyroscope bias and extrinsic rotation lets a VIO system recalibrate online after temperature or mechanical deformation changes the camera-IMU mounting, without an offline calibration tool.
  • The chi-square failure detector gives the system a practical signal to wait for more data, so a bad initialization is less likely to be accepted.
  • The MAP refinement across sliding windows makes the rotation and bias estimates improve continuously before translation arrives, giving the full VIO estimator a better warm start.

Reading between the lines

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

  • The paper does not prove observability; an analytic observability condition for the rotation-only cost would settle whether two keyframe pairs are always sufficient, and would show the motion profiles under which the six unknowns separate.
  • The weighting strategy is not tied to VIO: the same eigenvalue-based cost with per-feature covariance weighting could be applied to other geometric vision problems that minimize an eigenvalue, such as essential-matrix or homography estimation.
  • The failure-detection pass rate could be reused as a drift trigger: when the chi-square pass rate drops during operation, the system could re-run the rotation-only calibration rather than only using it at startup.
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

4 major / 7 minor

Summary. The paper proposes DOGE, a VIO initialization method that jointly estimates the camera-IMU extrinsic orientation and gyroscope bias using normal epipolar constraints (NEC). The central idea is to substitute a first-order model of the gyro-preintegrated rotation, parameterized by gyro-bias error δb_g and extrinsic-orientation error δθ_CI, into the NEC cost and minimize the minimum eigenvalue of the moment matrix M_ij. The method adds λ and feature-pair weighting, a Chi-square-based failure detector, and an IESKF/MAP refinement. Experiments on EuRoC and on a simulated pure-rotation scene compare favorably with VINS-Mono and Drt-VIO in accuracy, robustness, and runtime.

Significance. If the rotation-only solvability were established, the method would be a practically useful contribution to VIO initialization under loose or deformed extrinsics, because it would remove the requirement of translation parallax for rotational calibration and would lower initialization latency. The paper is commendable for deriving the coupled constraint, for the explicit weighting/outlier-rejection modules, and for documenting runtimes and failure-detection rates. However, the main theoretical claim is not supported as stated: the proposed cost has a symmetry (and hence a continuum of minima) when the motion is a pure rotation about a fixed axis, so the abstract's assertion that the problem 'can be solved in pure-rotation cases' needs qualification, and the two-keyframe-pair sufficiency argument is incomplete.

major comments (4)
  1. [Section III-A, Eqs. (8)-(12)] The rotation-only cost is not well posed for the pure-rotation cases claimed in the Abstract. Suppose all relative rotations are pure rotations about a fixed IMU axis n. Then every measured preintegrated rotation gamma_hat_ij and every correction Exp(J_b delta_b) (for delta_b parallel to n) is a rotation about n, so the predicted camera rotation X gamma_hat_ij Exp(J_b delta_b) X^{-1} is a rotation about the single camera axis q = R_CI n. Replacing X by X Exp(alpha n) leaves each predicted camera rotation exactly unchanged, because Exp(alpha n) commutes with gamma_hat_ij and with Exp(J_b delta_b). Hence the objective in Eq. (12) is constant along a one-parameter (and, if delta_b is also aligned with n, two-parameter) family of parameter values, lambda_min(M_ij) has a continuum of global minima, and the stacked Jacobian is rank-deficient. This is not an artifact of the first-order approximation in Eq. (9): it is an exact symmetry of the nonlinear cost. The paper should state an explicit rotational-excitation condition (for example, that the set of inter-frame rotation axes spans at least two dimensions) and should qualify the claim that the problem 'can be solved in pure-rotation cases.'
  2. [Section III-A, 'the size of K is at least 2'] The sufficiency argument for K >= 2 is incomplete. Even if the NEC provides three independent constraints per two-view setup for the rotation itself, the constraints in Eq. (12) are functions of the composite quantity X gamma_hat_ij Exp(J_b delta_b) X^{-1}; dependencies among these constraints as functions of the six unknowns are not ruled out. The fixed-axis example above shows that the stacked residual can have rank at most five (or lower) despite K >= 2. Please provide an observability/identifiability analysis of the stacked cost, or a numerical rank test of the Jacobian of Eq. (12), and state the conditions under which a unique global minimum exists.
  3. [Section III-A, Eq. (10)] The Jacobian J_theta_CI in Eq. (10) appears to be derived for a left perturbation of R_CI, whereas the manifold operators in Eq. (1) and the state update in Eq. (20) adopt the right-perturbation convention R ⊞ θ = R Exp(θ). Under the right-perturbation R_CI = bR_CI Exp(delta_theta_CI), the derivative of gamma_Ci_Cj with respect to delta_theta_CI is bR_CI (gamma_Ii_Ij^T - I), not bR_CI gamma_Ii_Ij^T bR_CI^T - I. Please clarify the perturbation convention used in the LM optimization and correct Eq. (10) if the right-perturbation convention is intended.
  4. [Section IV-B, Figs. 1 and 6] The simulation used to support the pure-rotation claim employs pure rotation with jointly varying roll, pitch, and yaw (Fig. 1), which provides the rotational excitation needed to avoid the fixed-axis degeneracy. It therefore does not validate the unqualified statement that the method works in pure-rotation cases. An additional experiment with rotation about a single fixed axis (e.g., a turntable motion) is needed; with the current formulation, the estimated extrinsic orientation should be non-identifiable in that scenario, confirming the need for an excitation condition.
minor comments (7)
  1. [Title and Abstract] 'An Extrinsic Orientation and Gyroscope Bias Estimation' is ungrammatical; consider 'Extrinsic Orientation and Gyroscope Bias Estimation' or 'A Method for Estimating Extrinsic Orientation and Gyroscope Bias.'
  2. [Section III-A, before Eq. (9)] The sentence introducing Eq. (9) is missing punctuation and the word 'as'; please revise for clarity.
  3. [Eq. (14)] The IMU preintegration covariance term carries a stray superscript k; the covariance of gamma_Ii_Ij does not depend on the point index.
  4. [Section IV-A] 'cause' should be 'because'; the caption of Fig. 2 also ends without a period.
  5. [Section IV-B] The sequence list 'V1 02, V1 03, V2 02, and V1 03' repeats V1 03; the last item is probably V2 03.
  6. [Section III-A] The sentence 'bγ_Ci_Cj represents the rotation estimation from frame Ci to frame Cj, which is also equal to bR_CiCj' introduces bR_CiCj before it is defined.
  7. [Fig. 4 caption] 'Ablation expriment' should be 'Ablation experiment.'

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the DOGE rotation-only cost is derived in-paper from standard NEC and IMU preintegration, and no fitted parameter is renamed as a prediction.

full rationale

The paper's central claim is that joint extrinsic orientation and gyroscope bias can be estimated from a rotation-only NEC residual. That residual is constructed explicitly in the paper: Eq. (9) propagates first-order errors through the preintegrated rotation, Eq. (11) builds M_ij from that perturbed rotation and the bearing vectors, and Eq. (12) minimizes the minimum eigenvalue. No target quantity from EuRoC or the simulated dataset is used to define the objective, and no constant appearing in the method is fitted to the reported errors. The hand-set values (0.5-pixel noise, Cauchy kernel, epsilon=0.8) affect robustness only and are not the quantities being predicted. The only substantive self-citation is reference [3], used for the downstream velocity and gravity stage after the extrinsic orientation and gyroscope bias have already been estimated; it is not load-bearing for the rotation-only derivation, which is self-contained. The observability concern raised for exact pure rotation about a fixed axis is a correctness or excitation-condition issue, not a circularity: even if Eq. (12) were degenerate in a particular motion regime, that would not make the derivation equivalent to its own inputs by construction. Since the derivation chain is independent of the experimental targets and the core estimator is not justified by a self-citation loop, the appropriate circularity finding is none.

Assumptions & free parameters 2 free parameters · 6 assumptions · 0 invented entities

The central claim rests on the standard IMU/preintegration model plus a set of assumptions about the NEC residual that are not proven: linearization validity at large extrinsic error, non-degeneracy under pure rotation, and the K>=2 constraint count. No new physical entities are postulated; the only hand-set constants are the failure detection threshold and assumed feature noise.

free parameters (2)
  • failure detection threshold epsilon = 0.8
    Hand-set pass rate threshold in Section III-A; if final-loop chi-square pass rate < 0.8 the initialization is rejected. Not fitted to data but tuned by the authors.
  • assumed feature observation noise std = 0.5 pixel
    Section III-A: 'we assume the observation noise of each point projection observes independent zero-mean Gauss distribution with 0.5-pixel standard deviation'. This sets all uncertainty weights and chi-square gating; it is an assumption, not measured.
assumptions (6)
  • domain assumption Standard IMU measurement model: gyro noise is Gaussian white noise n_g ~ N(0, sigma_g^2 I), and gyro bias evolves as a random walk with sigma_bg^2.
    Invoked in Section II-A, Eq. (2); this is the textbook IMU model from [13], not justified by data in the paper.
  • domain assumption First-order error-state approximation gamma = hat_gamma Exp(J_b delta_b + J_theta delta_theta) is accurate over the operating range of extrinsic errors up to 20 degrees.
    Eq. (9) in Section III-A; used to build the joint constraint. With large initial extrinsic error the linearization error is acknowledged, and the paper relies on iterative reweighting to compensate, but validity is not proven.
  • ad hoc to paper The minimum eigenvalue residual of the NEC moment matrix remains a valid and sufficiently informative cost for rotation estimation even when translation parallax is zero or near-zero (pure-rotation cases).
    Central to the 'rotation-only constraint' claim in Section I and Section III-A; the paper never proves observability or non-degeneracy at zero translation, where standard epipolar geometry degenerates.
  • domain assumption Deformation changes only the rotational part of the extrinsic parameters; the translational part changes at the millimeter level and can be ignored.
    Stated in Section I with reference [6]; the method does not estimate extrinsic translation, and all experiments inject only rotational errors.
  • domain assumption Gyroscope bias is constant within each sliding window during initialization.
    Stated in Section II-A; used to define the preintegration and the MAP prior in Eq. (19).
  • ad hoc to paper Each two-view setup with more than five co-visible points provides three independent constraints from NEC, so K>=2 is sufficient for the six unknowns.
    Claimed in Section III-A after Eq. (12), citing [11]; the counting is not derived in the paper and is questionable, see red flags.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DOGE: An Extrinsic Orientation and Gyroscope Bias Estimation for Visual-Inertial Odometry Initialization." pith.science (2026). https://pith.science/paper/TSXO2JK3

@misc{pith2026241208135,
  author       = {Pith},
  title        = {Pith review of: DOGE: An Extrinsic Orientation and Gyroscope Bias Estimation for Visual-Inertial Odometry Initialization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TSXO2JK3}},
  note         = {Machine review of arXiv:2412.08135}
}
read the original abstract

Most existing visual-inertial odometry (VIO) initialization methods rely on accurate pre-calibrated extrinsic parameters. However, during long-term use, irreversible structural deformation caused by temperature changes, mechanical squeezing, etc. will cause changes in extrinsic parameters, especially in the rotational part. Existing initialization methods that simultaneously estimate extrinsic parameters suffer from poor robustness, low precision, and long initialization latency due to the need for sufficient translational motion. To address these problems, we propose a novel VIO initialization method, which jointly considers extrinsic orientation and gyroscope bias within the normal epipolar constraints, achieving higher precision and better robustness without delayed rotational calibration. First, a rotation-only constraint is designed for extrinsic orientation and gyroscope bias estimation, which tightly couples gyroscope measurements and visual observations and can be solved in pure-rotation cases. Second, we propose a weighting strategy together with a failure detection strategy to enhance the precision and robustness of the estimator. Finally, we leverage Maximum A Posteriori to refine the results before enough translation parallax comes. Extensive experiments have demonstrated that our method outperforms the state-of-the-art methods in both accuracy and robustness while maintaining competitive efficiency.

Figures

Figures reproduced from arXiv: 2412.08135 by the authors.

Figure 1
Figure 1. Illustration of the advantage of the proposed method for AR applications: The top row of images shows the simulation scene and the motion trajectory of the camera. In the first 25 seconds, the camera has only pure rotational motion. The detailed configuration can be found in Sec. IV￾B. The bottom left image shows the estimation errors of rotational extrinsic parameters and gyroscope bias. It can be seen that our met… view at source ↗
Figure 3
Figure 3. Robustness evaluation: The top bar is colored according to the magnitude of the average angular velocity of the data segments (darker colors indicate higher magnitudes). The cyan points indicate data segments where a good estimation was obtained (b I g error is less than 50% and the RCI error is less than 5 degrees). The orange points indicate data segments where a bad estimation is detected. The red points indicate… view at source ↗
Figure 5
Figure 5. Test on EuRoC: Each methods adopt the same configuration as VINS-Mono. Due to ignoring gyroscope bias, VINS-Mono(Pre) performs worse than VINS-Mono(Post). On contrary, the proposed method gets an outstanding performance on almost all sequences. 0 5 10 15 20 25 Pure rotation duration (s) 0.0 0.1 0.2 0.3 0.4 0.5 Trajectory RMSE (m) VINS-Mono(Post) Drt-VIO VINS-Mono(Pre) Ours_woKF Ours [PITH_FULL_IMAGE:figures/full_fi… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Ablation expriment can be completed in less than 15ms, which is 4.3 times faster than VINS+VI-BA and 2.7 times faster than Drt+VI-BA. Ablation Experiment: We use the estimated error of RCiCj in degrees as the comprehensive evaluation indicator for the estimation of RCI…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

20 extracted references · 15 canonical work pages

  1. [1]

    Instant visual odometry initialization for mobile ar,

    A. Concha, M. Burri, J. Briales, C. Forster, and L. Oth, “Instant visual odometry initialization for mobile ar,” IEEE Transactions on Visualization and Computer Graphics, vol. 27, no. 11, pp. 4226–4235, 2021

    work page 2021

  2. [2]

    Orb-slam3: An accurate open-source library for visual, visual–inertial, and multimap slam,

    C. Campos, R. Elvira, J. J. G. Rodr ´ıguez, J. M. Montiel, and J. D. Tard ´os, “Orb-slam3: An accurate open-source library for visual, visual–inertial, and multimap slam,” IEEE Transactions on Robotics , vol. 37, no. 6, pp. 1874–1890, 2021

    2021

  3. [3]

    A rotation-translation-decoupled solution for robust and efficient visual-inertial initialization,

    Y . He, B. Xu, Z. Ouyang, and H. Li, “A rotation-translation-decoupled solution for robust and efficient visual-inertial initialization,” in Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 739–748, 2023

    work page 2023

  4. [4]

    Stereo-nec: Enhancing stereo visual-inertial slam initialization with normal epipolar constraints,

    W. Wang, C. Chou, G. Sevagamoorthy, K. Chen, Z. Chen, Z. Feng, Y . Xia, F. Cai, Y . Xu, and P. Mordohai, “Stereo-nec: Enhancing stereo visual-inertial slam initialization with normal epipolar constraints,” in IEEE International Conference on Robotics and Automation (ICRA) , pp. 2691–2697, 2024

    work page 2024

  5. [5]

    Unified temporal and spa- tial calibration for multi-sensor systems,

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

    work page 2013

  6. [6]

    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

  7. [7]

    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,” in IEEE Interna- tional Conference on Robotics and Automation (ICRA) , pp. 4666– 4672, IEEE, 2020

    work page 2020

  8. [8]

    Robust initialization of monocular visual-inertial estimation on aerial robots,

    T. Qin and S. Shen, “Robust initialization of monocular visual-inertial estimation on aerial robots,” in IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , pp. 4225–4232, IEEE, 2017

    work page 2017

Show all 20 references
  1. [9]

    Estimator initialization in vision- aided inertial navigation with unknown camera-imu calibration,

    T.-C. Dong-Si and A. I. Mourikis, “Estimator initialization in vision- aided inertial navigation with unknown camera-imu calibration,” in IEEE/RSJ International Conference on Intelligent Robots and Systems, pp. 1064–1071, IEEE, 2012

    2012

  2. [10]

    Online initialization and automatic camera-imu extrinsic calibration for monocular visual-inertial slam,

    W. Huang and H. Liu, “Online initialization and automatic camera-imu extrinsic calibration for monocular visual-inertial slam,” in IEEE Inter- national Conference on Robotics and Automation (ICRA) , pp. 5182– 5189, IEEE, 2018

    2018

  3. [11]

    Finding the exact rotation between two images independently of the translation,

    L. Kneip, R. Siegwart, and M. Pollefeys, “Finding the exact rotation between two images independently of the translation,” in in Springer European Conference on Computer Vision (ECCV) , pp. 696–709, Springer, 2012

    2012

  4. [12]

    Direct optimization of frame-to-frame rotation,

    L. Kneip and S. Lynen, “Direct optimization of frame-to-frame rotation,” in Proceedings of the IEEE International Conference on Computer Vision (ICCV) , pp. 2352–2359, 2013

    2013

  5. [13]

    Imu preinte- gration on manifold for efficient visual-inertial maximum-a-posteriori estimation,

    C. Forster, L. Carlone, F. Dellaert, and D. Scaramuzza, “Imu preinte- gration on manifold for efficient visual-inertial maximum-a-posteriori estimation,” in Robotics: Science and Systems XI , 2015

    2015

  6. [14]

    J. K. Uhlmann, Dynamic map building and localization: new theoret- ical foundations. PhD thesis, University of Oxford, 1995

    1995

  7. [15]

    The probabilistic normal epipolar constraint for frame-to-frame rotation optimization under uncertain feature positions,

    D. Muhle, L. Koestler, N. Demmel, F. Bernard, and D. Cremers, “The probabilistic normal epipolar constraint for frame-to-frame rotation optimization under uncertain feature positions,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR),...

    2022

  8. [16]

    Kalman filters on differentiable manifolds,

    D. He, W. Xu, and F. Zhang, “Kalman filters on differentiable manifolds,” arXiv preprint arXiv:2102.03804 , 2021

    2021 arXiv

  9. [17]

    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 , vol. 35, no. 10, pp. 1157–1163, 2016

    2016

  10. [18]

    An iterative image registration technique with an application to stereo vision,

    B. D. Lucas and T. Kanade, “An iterative image registration technique with an application to stereo vision,” in 7th international joint confer- ence on Artificial intelligence (IJCAI) , vol. 2, pp. 674–679, 1981

    1981

  11. [19]

    Good features to track,

    J. Shi et al. , “Good features to track,” in Proceedings of IEEE con- ference on computer vision and pattern recognition (CVPR) , pp. 593– 600, IEEE, 1994

    1994

  12. [20]

    Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography,

    M. A. Fischler and R. C. Bolles, “Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography,” Commun. ACM, vol. 24, no. 6, pp. 381–395, 1981

    1981

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.