Pith. sign in

REVIEW 2 major objections 5 minor 69 references

SmartPoser: Arm Pose Estimation with a Smartphone and Smartwatch Using UWB and IMU Data

T0 review · 2 major / 5 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read SmartPoser claims that an off-the-shelf smartphone and smartwatch, fused through ultra-wideband distance and inertial data, can track elbow and wrist positions with a median error of 11.0 cm in real time, with no user training data.

desk verdict Solid commodity-device pose tracking with a credible 11-cm claim, but the 'first UWB+IMU' framing overreaches and the headline number depends on an unstated arm-span input. read the letter →

arxiv 2509.03451 v1 pith:GRCRZP33 submitted 2025-09-03 cs.HC cs.CVcs.GRcs.RO

Nathan DeVrio , Vimal Mollyn , Chris Harrison This is my paper · ORCID
classification cs.HCcs.CVcs.GRcs.RO
keywords armposeestimationultra-widebandrangingIMUfusionsmartwatchsensingsmartphonewearablemotiontrackingreal-time
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

Arm pose tracking typically needs cameras, body-worn sensor suits, or instrumented rooms — none of which match what consumers own. SmartPoser claims that two devices people already carry — a smartphone in the front pocket and a smartwatch on the wrist — can do it together. The trick is pairing a measurement that drifts (inertial orientation and acceleration from each device's IMU) with one that does not (ultra-wideband radio measures the absolute distance between the two devices about five times per second). Two small recurrent networks clean up the noisy distance signal and convert the combined stream into 3D elbow and wrist positions, normalized to the wearer's arm span. In a 10-person study with leave-one-participant-out evaluation, the system reports a median error of 11.0 cm averaged over elbow and wrist — the paper argues this puts commodity phone-and-watch tracking on par with far more instrumented systems.

What carries the argument

Two complementary measurements carry the argument: UWB ranging (~5 Hz) gives an absolute, drift-free distance between phone and watch; each device's IMU (~25 Hz) gives orientation and acceleration. Two small bidirectional LSTM networks do the computation — 46.4k parameters total. The UWB correction network cleans the raw distance stream using IMU context, cutting mean distance error from 21.0 cm to 8.6 cm. The pose estimator then converts corrected distance plus IMU features into 3D shoulder, elbow, and wrist positions normalized to arm span, reading the n−5th frame (200 ms behind real time) as its output. Side-by-side devices plus a T-pose align the two reference frames.

What would settle it

Run SmartPoser on participants whose arm span is deliberately mis-entered by ±10% and ±20%, and who rotate the phone in the pocket after calibration, then compare median elbow/wrist error against marker-based ground truth. If the 11.0 cm figure holds only with an exact arm span and an undisturbed calibration, the accuracy claim overstates robustness to inputs a real user controls.

Watch

Extended reading notes

Core claim

SmartPoser's central claim is that the absolute distance between a smartphone and a smartwatch — measured with UWB time-of-flight ranging — is the missing signal that makes arm pose estimation practical on consumer hardware. Inertial data drifts and is relative; UWB distance is absolute and stable, so fusing the two gives the pose estimator an anchor. A UWB correction network reduces raw distance error from a mean of 21.0 cm to 8.6 cm, and a pose estimator outputs shoulder, elbow, and wrist positions normalized to the user's arm span, then reprojected to centimeters by that span. With leave-one-participant-out cross-validation against Kinect-derived ground truth, the system reports median er

Load-bearing premise

The 11.0 cm median assumes the user's arm span is known and the two-step calibration (devices held side by side, then a T-pose) was performed correctly; the paper does not quantify how much error grows when the arm span is wrong or the phone shifts or rotates in the pocket afterward.

Editorial extensions

If this is right

  • Arm tracking becomes a software feature: UWB-capable phone-and-watch pairs could run pose estimation as an app update, with no cameras, body suits, or external infrastructure.
  • Fitness, rehabilitation, life-logging, occupational-safety, and context-aware assistant applications gain a continuous, real-time arm-position stream in mobile settings where prior systems required static or heavily instrumented setups.
  • The reported accuracy (8.7 cm elbow, 13.3 cm wrist, 11.0 cm pooled median) places commodity phone-and-watch tracking in the same range as far more instrumented systems, including ArmTrak, MUSE, and IMUPoser.
  • The approach is claimed to keep working while the user moves — a capability ArmTrak lacked — and the same pipeline is expected to extend to other phone placements such as rear pockets and jacket pockets, and to other UWB devices like earbuds or smart glasses.

Reading between the lines

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

  • Because every reported coordinate is normalized to arm span and then scaled back, an error in the user's arm-span estimate propagates roughly linearly into absolute position; the paper never measures this sensitivity, so everyday accuracy could deviate from 11.0 cm whenever a user's span is misentered.
  • The paper's own heatmaps show the worst errors where the UWB signal is blocked by the body (above the shoulder, at the right thigh); a second UWB anchor — e.g., earbuds — or a stronger correction model would likely attack the worst cases more than the median.
  • The same corrector-plus-regressor recipe could be retrained for other body parts: a pocket phone and a UWB-equipped shoe might track leg pose, or a UWB earbud could replace the phone as the absolute reference for the watch.
  • The fixed 200 ms output lag sets a floor on interactivity; gesture-input and gaming applications that need immediate response would have to trade accuracy for a shorter context window or a shallower network.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 5 minor

Summary. SmartPoser proposes a software-only arm-pose tracking system using an off-the-shelf smartphone and smartwatch. It fuses UWB distance measurements between the two devices with their IMU orientations and accelerations, feeding a two-stage bidirectional-LSTM pipeline: a UWB correction network that cleans noisy body-occluded range measurements, followed by a pose estimator that outputs shoulder, elbow, and wrist positions in a normalized arm-span coordinate frame. The system is trained end-to-end on Kinect-derived ground truth from 10 participants and evaluated with leave-one-participant-out cross-validation. The headline result is a median elbow/wrist positional error of 11.0 cm (SD=0.9 cm), with an ablation showing UWB reduces error by 20--25% relative to IMU-only input. The authors also report a real-time 25 Hz implementation and release the dataset, models, and code.

Significance. If the 11.0 cm median error is reproducible in the claimed deployment conditions, this is a meaningful advance: it would be the first demonstration of arm-pose tracking at this accuracy using only a phone and a watch in normal wearing positions, with no per-user training data and no external infrastructure. The paper's strengths include a real user study with 280k frames, leave-one-participant-out evaluation, a clean UWB/IMU ablation, an on-device CoreML latency measurement, and open-sourced data and models. These are concrete and valuable contributions beyond the headline number. However, the absolute error depends on an external input—the user's arm span—whose acquisition and sensitivity are not described, and the comparative claims against prior systems rest on incompatible datasets. Both issues are addressable but need attention before the central claim is fully supported.

major comments (2)
  1. [Sec. 3.3, 3.4, and 8] The reported 11.0 cm median error is not an end-to-end system result as stated because the model's input features and output coordinates are both scaled by the user's arm span, yet the paper never specifies how arm span is obtained in practice or how sensitive the result is to that input. Section 3.3 says UWB distances are 'scaled by the wearer's arm span' and Section 3.4 says joint locations are 'normalized to the user's arm span' and then reprojected into real-world units by multiplying by arm span. A 10% error in arm span will propagate directly into the output scale and also alter the UWB-correction input features, so the measured 11.0 cm could change by several centimeters. Section 8 lists calibration and pocket placement as limitations but is silent on arm-span acquisition. The authors should state how arm span is measured or supplied (e.g., user entry, height-based estimate, or a
  2. [Sec. 5.5] The comparison to ArmTrak, IMUPoser, and MUSE is made across datasets with different tasks, protocols, and evaluation metrics. The paper acknowledges 'these systems were evaluated on different datasets' but then asserts the comparison is applicable. This does not invalidate the absolute 11.0 cm measurement, but it does not support the stronger framing in the introduction that SmartPoser achieves 'the best tracking accuracy among systems that do not require special user instrumentation or external infrastructure.' A direct comparison on a shared benchmark or a reimplementation is needed, or the claims should be softened to 'comparable in magnitude to previously reported results under different evaluation conditions.'
minor comments (5)
  1. [Sec. 5.5] The sentence 'our system achieved a real-time (25 Hz) median error of 11.0 cm for the shoulder and wrist' appears to be a typo: the headline metric is for elbow and wrist, and the next clause says '13.3 cm for just the wrist.' Please correct.
  2. [Sec. 3.4, 5.5] The 'real-time' description should be qualified by the 200 ms output delay from using the n-5th frame. This is disclosed in Section 3.4 but not in the results summary or comparison section; a reader could interpret '25 Hz real-time' as zero-latency. Please state this latency explicitly wherever real-time performance is claimed.
  3. [Sec. 5.3 / Fig. 7] The text says the scatter plot uses 'all 131k frames of study data from our gross arm pose procedure,' while Section 4.3 reports 129K frames for gross motions and Figure 7's caption says 280k datapoints. These numbers should be reconciled.
  4. [Sec. 4.3] The Kinect pairing procedure is described as 'paired it with the most recent Kinect frame' but the Kinect frame rate is not stated. Please report the Kinect capture rate so readers can assess synchronization latency, particularly for fast arm motions.
  5. [Sec. 5.3] The caption of Figure 7 states R^2 = 0.72 for the corrected-UWB vs Kinect-distance fit, but the text in Section 5.3 does not interpret this value in context. It would help to state what fraction of variance remains unexplained and how that relates to the residual 8.6 cm error.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the headline 11.0 cm error is a held-out empirical measurement against external Kinect ground truth; arm-span normalization is a linear scaling, not a self-referential quantity; minor self-citations are not load-bearing.

full rationale

SmartPoser's central claim—a median positional error of 11.0 cm using only a phone and watch—is an empirical result, not a derivation that reduces to its own inputs. The two networks (UWB correction and pose estimation) are trained against Azure Kinect-derived distances and joint positions (Secs. 3.3, 3.5) and evaluated with leave-one-participant-out cross-validation (Sec. 5.1), so the reported error is measured on participants and frames never seen in training. The paper even reports a training-set wrist error of 11.9 cm versus 13.3 cm cross-validated (Sec. 5.2), showing no memorization. The arm-span normalization (Secs. 3.3–3.4) scales both inputs and outputs by a constant body measurement; it is a linear unit conversion, not a fitted parameter renamed as a prediction. However, the paper never states how arm span is obtained in practice and provides no sensitivity analysis (Sec. 8 discusses calibration but is silent on arm span); this is a reproducibility/robustness gap, not a circularity. Self-citations appear in related work and example uses (IMUPoser [32], Harrison & Hudson [13], Laput & Harrison [23, 24], Ubicoustics [22]), and IMUPoser is listed alongside external works (TransPose, DIP, PIP) for the T-pose calibration procedure; that procedure's validity is confirmed by the empirical study itself, so no load-bearing argument depends on an unverified self-citation. No equation in the paper is defined in terms of the quantity it purports to predict.

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

The central claim rests on the learned RNN weights (fitted to Kinect ground truth) and on user-specific calibration inputs: arm span and the side-by-side/T-pose reference alignment. The free parameters listed are the user-dependent or hand-chosen constants that convert the model's normalized output into the reported centimeters; the network weights themselves are not listed separately because they are the fitted model rather than a number introduced ad hoc.

free parameters (4)
  • user_arm_span
    Used to scale UWB distances and to reproject normalized joint coordinates into centimeters (Section 3.4); the paper does not state how it is obtained, yet the 11.0 cm error is directly scale-dependent.
  • prediction_output_delay = 5 frames (200 ms)
    The model outputs the n-5th frame instead of the current frame to reduce error, trading 200 ms latency for accuracy (Section 3.4). This delay is included in the reported real-time error.
  • acceleration_scaling_factor = 30
    Acceleration inputs are scaled by a factor of 30 to aid training, following prior work [58,59] (Section 3.3).
  • network_hyperparameters = hidden sizes 8/32, window 125, lr 3e-4
    Architecture choices (LSTM hidden units, 5 s window, Adam learning rate) follow prior work or were set by hand and not systematically ablated.
assumptions (4)
  • domain assumption Azure Kinect skeleton output is accurate enough to serve as ground truth for training and evaluation.
    All training and error numbers are computed against Kinect-derived joint positions; any systematic Kinect error appears directly in the reported errors.
  • domain assumption The UWB distance measurement and its noise can be corrected using IMU orientation and acceleration data.
    The UWB Correction RNN assumes the structured body-occlusion error in UWB ranging is learnable from the IMU streams (Section 3.3).
  • domain assumption The phone remains fixed in the front pocket and the watch on the left wrist, preserving the calibration.
    The system assumes device placements stay constant; Section 8 limits the method to pocket placement.
  • domain assumption IMU orientation from Core Motion provides a consistent relative frame after the side-by-side and T-pose calibration.
    The alignment procedure in Section 4 is assumed accurate and stable over time.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SmartPoser: Arm Pose Estimation with a Smartphone and Smartwatch Using UWB and IMU Data." pith.science (2026). https://pith.science/paper/GRCRZP33

@misc{pith2026250903451,
  author       = {Pith},
  title        = {Pith review of: SmartPoser: Arm Pose Estimation with a Smartphone and Smartwatch Using UWB and IMU Data},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GRCRZP33}},
  note         = {Machine review of arXiv:2509.03451}
}
read the original abstract

The ability to track a user's arm pose could be valuable in a wide range of applications, including fitness, rehabilitation, augmented reality input, life logging, and context-aware assistants. Unfortunately, this capability is not readily available to consumers. Systems either require cameras, which carry privacy issues, or utilize multiple worn IMUs or markers. In this work, we describe how an off-the-shelf smartphone and smartwatch can work together to accurately estimate arm pose. Moving beyond prior work, we take advantage of more recent ultra-wideband (UWB) functionality on these devices to capture absolute distance between the two devices. This measurement is the perfect complement to inertial data, which is relative and suffers from drift. We quantify the performance of our software-only approach using off-the-shelf devices, showing it can estimate the wrist and elbow joints with a \hl{median positional error of 11.0~cm}, without the user having to provide training data.

Figures

Figures reproduced from arXiv: 2509.03451 by the authors.

Figure 1
Figure 1. SmartPoser fuses UWB distance and IMU data from an off-the-shelf smartphone and smartwatch (A) to track a [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. SmartPoser system overview. Our model has 46.4k [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Example user motion sequence over one minute. Raw UWB distance is plotted in blue — note the significant errors, [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Example set of predicted poses. Vertex error (vs. [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 5
Figure 5. Figure 5: The 20 terminal arm poses utilized in our user study. Note that participants were asked to transition between these [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: Median per-joint positional error was 11.0 cm av [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 7
Figure 7. Figure 7: Left: Plot of corrected UWB distance vs. Kinect [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]
Figure 9
Figure 9. Figure 9: Three rows of heatmaps visualizing maximum ab [PITH_FULL_IMAGE:figures/full_fig_p007_9.png]
Figure 10
Figure 10. Figure 10: Ablation results showing the effectiveness using [PITH_FULL_IMAGE:figures/full_fig_p008_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

69 extracted references · 53 canonical work pages

  1. [1]

    Apple. 2022. Core Motion | Apple Developer Documentation. https://developer. apple.com/documentation/coremotion/

    work page 2022

  2. [2]

    Apple. 2022. Nearby Interaction | Apple Developer Documentation. https: //developer.apple.com/documentation/nearbyinteraction

    work page 2022

  3. [3]

    Apple. 2023. Set up Handwashing on Apple Watch. https://support.apple.com/ guide/watch/set-up-handwashing-apdc9b9f04a8/watchos

    work page 2023

  4. [4]

    Bachmann, Robert B

    Eric R. Bachmann, Robert B. McGhee, Xiaoping Yun, and Michael J. Zyda. 2001. Inertial and magnetic posture tracking for inserting humans into networked virtual environments. In Proceedings of the ACM symposium on Virtual reality software and technology (VRST ’01) . Association for Computing Machinery, New York, NY, USA, 9–16. https://doi.org/10.1145/505008.505011

    work page arXiv 2001

  5. [5]

    Alejandro Baldominos, Yago Saez, and Cristina García del Pozo. 2015. An Ap- proach to Physical Rehabilitation Using State-of-the-art Virtual Reality and Motion Tracking Technologies. Procedia Computer Science 64 (2015), 10–16. https://doi.org/10.1016/j.procs.2015.08.457

    work page doi:10.1016/j.procs.2015.08.457 2015

  6. [6]

    Ling Bao and Stephen S. Intille. 2004. Activity Recognition from User-Annotated Acceleration Data. In Pervasive Computing (Lecture Notes in Computer Science) , Alois Ferscha and Friedemann Mattern (Eds.). Springer, Berlin, Heidelberg, 1–17. https://doi.org/10.1007/978-3-540-24646-6_1

    doi:10.1007/978-3-540-24646-6_1 2004

  7. [7]

    Frédéric Bosché, Mohamed Abdel-Wahab, and Ludovico Carozza. 2016. Towards a Mixed Reality System for Construction Trade Training.Journal of Computing in Civil Engineering 30, 2 (March 2016), 04015016. https://doi.org/10.1061/(ASCE) CP.1943-5487.0000479 Publisher: American Society of Civil Engineers

    arXiv 2016

  8. [8]

    Sanchez, Pramod Murthy, and Didier Stricker

    Hammad Tanveer Butt, Bertram Taetz, Mathias Musahl, Maria A. Sanchez, Pramod Murthy, and Didier Stricker. 2021. Magnetometer Robust Deep Hu- man Pose Regression With Uncertainty Prediction Using Sparse Body Worn Magnetic Inertial Measurement Units. IEEE Access 9 (2021), 36657–36673. https://doi.org/10.1109/ACCESS.2021.3062545 Conference Name: IEEE Access

    work page arXiv 2021

Show all 69 references
  1. [9]

    J. A. Corrales, F. A. Candelas, and F. Torres. 2008. Hybrid tracking of human operators using IMU/UWB data fusion by a Kalman filter. In 2008 3rd ACM/IEEE International Conference on Human-Robot Interaction (HRI) . 193–200. https: //doi.org/10.1145/1349822.1349848 ISSN: 2167-2148

    2008

  2. [10]

    Han Ding, Jinsong Han, Longfei Shangguan, Wei Xi, Zhiping Jiang, Zheng Yang, Zimu Zhou, Panlong Yang, and Jizhong Zhao. 2017. A Platform for Free-Weight Exercise Monitoring with Passive Tags. IEEE Transactions on Mobile Comput- ing 16, 12 (Dec. 2017), 3279–3293. https://doi.or...

    2017

  3. [11]

    Google. 2022. Android Developers, Ultra-wideband (UWB) communication. https://developer.android.com/guide/topics/connectivity/uwb

    2022

  4. [12]

    Bernhard Großwindhager, Carlo Alberto Boano, Michael Rath, and Kay Römer

  5. [13]

    Chris Harrison and Scott E. Hudson. 2008. Lightweight material detection for placement-aware mobile computing. In Proceedings of the 21st annual ACM sym- posium on User interface software and technology (UIST ’08) . Association for Computing Machinery, New York, NY, USA, 279–2...

    2008

  6. [14]

    P Helliwell and W Taylor. 2004. Repetitive strain injury. Postgraduate Medical Journal 80, 946 (Aug. 2004), 438–443. https://doi.org/10.1136/pgmj.2003.012591

    2004 arXiv

  7. [15]

    Hol, Fred Dijkstra, Henk Luinge, and Thomas B

    Jeroen D. Hol, Fred Dijkstra, Henk Luinge, and Thomas B. Schon. 2009. Tightly coupled UWB/IMU pose estimation. In 2009 IEEE International Conference on Ultra-Wideband. 688–692. https://doi.org/10.1109/ICUWB.2009.5288724

    2009

  8. [16]

    Black, Otmar Hilliges, and Gerard Pons-Moll

    Yinghao Huang, Manuel Kaufmann, Emre Aksan, Michael J. Black, Otmar Hilliges, and Gerard Pons-Moll. 2018. Deep inertial poser: learning to reconstruct human pose from sparse inertial measurements in real time. ACM Transactions on Graphics 37, 6 (Dec. 2018), 1–15. https://doi.o...

    2018

  9. [17]

    Grignani

    Fumiko Ichikawa, Jan Chipchase, and R. Grignani. 2005. Where’s the phone? A study of mobile phone location in public spaces. Mobile Technology, Applications and Systems, 2005 2nd International Conference on (Jan. 2005), 1–8

    2005

  10. [18]

    Intille, Ling Bao, Emmanuel Munguia Tapia, and John Rondoni

    Stephen S. Intille, Ling Bao, Emmanuel Munguia Tapia, and John Rondoni. 2004. Acquiring in situ training data for context-aware ubiquitous computing applica- tions. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI ’04). Association for Computi...

    2004

  11. [19]

    Silje Ekroll Jahren, Niels Aakvaag, Frode Strisland, Andreas Vogl, Alessandro Lib- erale, and Anders E. Liverud. 2021. Towards Human Motion Tracking Enhanced by Semi-Continuous Ultrasonic Time-of-Flight Measurements. Sensors 21, 7 (Jan. 2021), 2259. https://doi.org/10.3390/s21072259

    2021 doi

  12. [20]

    Jiaxi Jiang, Paul Streli, Huajian Qiu, Andreas Fender, Larissa Laich, Patrick Snape, and Christian Holz. 2022. AvatarPoser: Articulated Full-Body Pose Tracking from Sparse Motion Sensing. http://arxiv.org/abs/2207.13784 arXiv:2207.13784 [cs]

    2022 arXiv

  13. [21]

    Haojian Jin, Zhijian Yang, Swarun Kumar, and Jason I. Hong. 2018. Towards Wearable Everyday Body-Frame Tracking using Passive RFIDs. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies 1, 4 (Jan. 2018), 145:1–145:23. https://doi.org/10.1145/3161199

    2018 doi

  14. [22]

    Gierad Laput, Karan Ahuja, Mayank Goel, and Chris Harrison. 2018. Ubicoustics: Plug-and-Play Acoustic Activity Recognition. In Proceedings of the 31st Annual ACM Symposium on User Interface Software and Technology (UIST ’18). Association for Computing Machinery, New York, NY, ...

    2018

  15. [23]

    Gierad Laput and Chris Harrison. 2019. Sensing Fine-Grained Hand Activity with Smartwatches. In Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems (CHI ’19) . Association for Computing Machinery, New York, NY, USA, 1–13. https://doi.org/10.1145/329060...

    2019

  16. [24]

    Gierad Laput, Robert Xiao, and Chris Harrison. 2016. ViBand: High-Fidelity Bio- Acoustic Sensing Using Commodity Smartwatch Accelerometers. In Proceedings of the 29th Annual Symposium on User Interface Software and Technology (UIST ’16). Association for Computing Machinery, Ne...

    2016

  17. [25]

    Tommaso Lisini Baldi, Francesco Farina, Andrea Garulli, Antonio Giannitrapani, and Domenico Prattichizzo. 2020. Upper Body Pose Estimation Using Wearable Inertial Sensors and Multiplicative Kalman Filter. IEEE Sensors Journal 20, 1 (Jan. 2020), 492–500. https://doi.org/10.1109...

    2020

  18. [26]

    Huajun Liu, Xiaolin Wei, Jinxiang Chai, Inwoo Ha, and Taehyun Rhee. 2011. Realtime human motion control with a small number of inertial sensors. In Symposium on Interactive 3D Graphics and Games (I3D ’11) . Association for Computing Machinery, New York, NY, USA, 133–140. https...

    2011

  19. [27]

    Shi Qiang Liu, Jun Chang Zhang, and Rong Zhu. 2020. A Wearable Human Motion Tracking Device Using Micro Flow Sensor Incorporating a Micro Accelerometer. IEEE Transactions on Biomedical Engineering 67, 4 (April 2020), 940–948. https: //doi.org/10.1109/TBME.2019.2924689 Conferen...

    2020

  20. [28]

    Marins, Xiaoping Yun, E.R

    J.L. Marins, Xiaoping Yun, E.R. Bachmann, R.B. McGhee, and M.J. Zyda. 2001. An extended Kalman filter for quaternion-based orientation estimation using MARG sensors. In Proceedings 2001 IEEE/RSJ International Conference on Intelligent Robots and Systems. Expanding the Societal...

    2001

  21. [29]

    Mekonnen, Eric Slottke, Heinrich Luecken, Christoph Steiner, and Armin Wittneben

    Zemene W. Mekonnen, Eric Slottke, Heinrich Luecken, Christoph Steiner, and Armin Wittneben. 2010. Constrained maximum likelihood positioning for UWB based human motion tracking. In 2010 International Conference on Indoor Posi- tioning and Indoor Navigation . 1–10. https://doi....

    2010

  22. [30]

    Jochen Meyer, Jutta Fortmann, Merlin Wasmann, and Wilko Heuten. 2015. Mak- ing Lifelogging Usable: Design Guidelines for Activity Trackers. In MultiMedia Modeling (Lecture Notes in Computer Science) , Xiangjian He, Suhuai Luo, Dacheng Tao, Changsheng Xu, Jie Yang, and Muhammad...

    2015 doi

  23. [31]

    Markus Miezal, Gabriele Bleser, Norbert Schmitz, and Didier Stricker. 2013. A generic approach to inertial tracking of arbitrary kinematic chains. InProceedings of the 8th International Conference on Body Area Networks (BodyNets ’13) . ICST (Institute for Computer Sciences, So...

    2013 doi

  24. [32]

    Vimal Mollyn, Riku Arakawa, Mayank Goel, Chris Harrison, and Karan Ahuja

  25. [33]

    Kim Doang Nguyen, I-Ming Chen, Zhiqiang Luo, Song Huat Yeo, and Henry Been-Lirn Duh. 2011. A Wearable Sensing System for Tracking and Monitoring of Functional Arm Movement. IEEE/ASME Transactions on Mechatronics 16, 2 (April 2011), 213–220. https://doi.org/10.1109/TMECH.2009.2...

    2011

  26. [34]

    Vijay Rajanna, Raniero Lara-Garduno, Dev Jyoti Behera, Karthic Madanagopal, Daniel Goldberg, and Tracy Hammond. 2014. Step up life: a context aware health assistant. In Proceedings of the Third ACM SIGSPATIAL International Workshop on the Use of GIS in Public Health (HealthGIS...

    2014

  27. [35]

    Mary Redmayne. 2017. Where’s Your Phone? A Survey of Where Women Aged 15-40 Carry Their Smartphone and Related Risk Perception: A Survey and Pilot Study. PLoS ONE 12, 1 (Jan. 2017), e0167996. https://doi.org/10.1371/journal. pone.0167996

    2017 doi

  28. [36]

    Yili Ren, Zi Wang, Sheng Tan, Yingying Chen, and Jie Yang. 2022. Winect: 3D Human Pose Tracking for Free-form Activity Using Commodity WiFi.Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies 5, 4 (Dec. 2022), 176:1–176:29. https://doi.org/10.11...

    2022 doi

  29. [37]

    Yili Ren, Zi Wang, Yichao Wang, Sheng Tan, Yingying Chen, and Jie Yang. 2022. GoPose: 3D Human Pose Estimation Using WiFi. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies 6, 2 (July 2022), 69:1– 69:25. https://doi.org/10.1145/3534605

    2022 doi

  30. [38]

    Daniel Roetenberg, Henk Luinge, and Per Slycke. 2013. Xsens MVN: Full 6DOF Human Motion Tracking Using Miniature Inertial Sensors. (2013), 9

    2013

  31. [39]

    Slycke, and Peter H

    Daniel Roetenberg, Per J. Slycke, and Peter H. Veltink. 2007. Ambulatory Po- sition and Orientation Tracking Fusing Magnetic and Inertial Sensing. IEEE Transactions on Biomedical Engineering 54, 5 (May 2007), 883–890. https: //doi.org/10.1109/TBME.2006.889184 Conference Name: ...

    2007

  32. [40]

    Arindam Sengupta, Feng Jin, Renyuan Zhang, and Siyang Cao. 2020. mm-Pose: Real-Time Human Skeletal Posture Estimation Using mmWave Radars and CNNs. IEEE Sensors Journal 20, 17 (Sept. 2020), 10032–10044. https://doi.org/10.1109/ JSEN.2020.2991741 Conference Name: IEEE Sensors Journal

    2020

  33. [41]

    Sheng Shen, Mahanth Gowda, and Romit Roy Choudhury. 2018. Closing the Gaps in Inertial Motion Tracking. In Proceedings of the 24th Annual International Conference on Mobile Computing and Networking (New Delhi, India) (MobiCom ’18). Association for Computing Machinery, New York...

    2018

  34. [42]

    Sheng Shen, He Wang, and Romit Roy Choudhury. 2016. I am a Smartwatch and I can Track my User’s Arm. In Proceedings of the 14th Annual International Conference on Mobile Systems, Applications, and Services (MobiSys ’16). Association for Computing Machinery, New York, NY, USA, ...

    2016

  35. [43]

    Shixin (Cindy) Shen and Ronald A. House. 2017. Hand-arm vibration syndrome. Canadian Family Physician 63, 3 (March 2017), 206–210. https://www.ncbi.nlm. nih.gov/pmc/articles/PMC5349719/

    2017

  36. [44]

    Yongkun Song, Tian Jin, Yongpeng Dai, Yongping Song, and Xiaolong Zhou. 2021. Through-Wall Human Pose Reconstruction via UWB MIMO Radar and 3D CNN. Remote Sensing 13, 2 (Jan. 2021), 241. https://doi.org/10.3390/rs13020241

    2021 doi

  37. [45]

    Elizabeth Spencer. 2020. The Pocket: A Hidden History of Women’s Lives, 1660-1900. Cultural and Social History 17, 4 (Aug. 2020), 570–572. https://doi.org/10.1080/14780038.2020.1810955 Publisher: Routledge _eprint: https://doi.org/10.1080/14780038.2020.1810955

    2020

  38. [46]

    tesych. 2020. About Azure Kinect DK. https://docs.microsoft.com/en-us/azure/ kinect-dk/about-azure-kinect-dk

    2020

  39. [47]

    Daniel Vlasic, Rolf Adelsberger, Giovanni Vannucci, John Barnwell, Markus Gross, Wojciech Matusik, and Jovan Popović. 2007. Practical motion capture in everyday surroundings. ACM Transactions on Graphics 26, 3 (July 2007), 35. https://doi.org/10.1145/1276377.1276421

    2007

  40. [48]

    Black, and Gerard Pons-Moll

    Timo von Marcard, Bodo Rosenhahn, Michael J. Black, and Gerard Pons-Moll

  41. [49]

    Chuyu Wang, Jian Liu, Yingying Chen, Lei Xie, Hong Bo Liu, and Sanclu Lu

  42. [50]

    Wenchuan Wei, Keiko Kurita, Jilong Kuang, and Alex Gao. 2021. Real-Time 3D Arm Motion Tracking Using the 6-axis IMU Sensor of a Smartwatch. In2021 IEEE 17th International Conference on Wearable and Implantable Body Sensor Networks (BSN). 1–4. https://doi.org/10.1109/BSN51625.2...

    2021

  43. [51]

    Scott Saponas, and A.J

    Jason Wiese, T. Scott Saponas, and A.J. Bernheim Brush. 2013. Phoneprioception: enabling mobile phones to infer where they are kept. InProceedings of the SIGCHI Conference on Human Factors in Computing Systems . ACM, Paris France, 2157–

    2013

  44. [52]

    Wikipedia. 2022. List of best-selling video games. https://en.wikipedia.org/ w/index.php?title=List_of_best-selling_video_games&oldid=1104319901 Page Version ID: 1104319901

    2022

  45. [53]

    RF-Kinect: A Wearable RFID-based Approach Towards 3D Body Movement Tracking. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous 10 SmartPoser: Arm Pose Estimation with a Smartphone and Smartwatch Using UWB and IMU Data UIST ’23, October 29-November 1, 2023,...

    2023 doi

  46. [54]

    Wilk, Michael Walsh, and Brendan O’Flynn

    Mariusz P. Wilk, Michael Walsh, and Brendan O’Flynn. 2021. Multimodal Sensor Fusion for Low-Power Wearable Human Motion Tracking Systems in Sports Applications. IEEE Sensors Journal 21, 4 (Feb. 2021), 5195–5212. https://doi.org/ 10.1109/JSEN.2020.3030779 Conference Name: IEEE ...

    2021

  47. [55]

    Frieder Wittmann, Olivier Lambercy, and Roger Gassert. 2019. Magnetometer- Based Drift Correction During Rest in IMU Arm Motion Tracking. Sensors 19, 6 (Jan. 2019), 1312. https://doi.org/10.3390/s19061312

    2019 doi

  48. [56]

    Xuesu Xiao and Shuayb Zarar. 2018. A Wearable System for Articulated Human Pose Tracking Under Uncertainty of Sensor Placement. In 2018 7th IEEE Interna- tional Conference on Biomedical Robotics and Biomechatronics (Biorob) . 1144–1150. https://doi.org/10.1109/BIOROB.2018.8487...

    2018

  49. [57]

    Chao Yang, Lingxiao Wang, Xuyu Wang, and Shiwen Mao. 2022. Environment Adaptive RFID-Based 3D Human Pose Tracking With a Meta-Learning Approach. IEEE Journal of Radio Frequency Identification 6 (2022), 413–425. https://doi.org/ 10.1109/JRFID.2022.3140256 Conference Name: IEEE ...

    2022

  50. [58]

    Wikipedia. 2022. List of UWB-enabled mobile devices. https://en.wikipedia.org/ w/index.php?title=List_of_UWB-enabled_mobile_devices&oldid=1109918576 Page Version ID: 1109918576

    2022

  51. [59]

    Xinyu Yi, Yuxiao Zhou, and Feng Xu. 2021. TransPose: real-time 3D human translation and pose estimation with six inertial sensors. ACM Transactions on Graphics 40, 4 (July 2021), 86:1–86:13. https://doi.org/10.1145/3450626.3459786

    2021

  52. [60]

    Alexander David Young. 2010. Use of Body Model Constraints to Improve Accuracy of Inertial Motion Capture. In 2010 International Conference on Body Sensor Networks. 180–186. https://doi.org/10.1109/BSN.2010.30 ISSN: 2376-8894

    2010 doi

  53. [61]

    Han Zhou, Yi Gao, Xinyi Song, Wenxin Liu, and Wei Dong. 2020. LimbMotion: Decimeter-Level Limb Tracking for Wearable-Based Human-Computer Interac- tion. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol. 3, 4, Article 161 (sep 2020), 24 pages. https://doi.org/10.1145/3369836

    2020 doi

  54. [62]

    Huiyu Zhou, Thomas Stone, Huosheng Hu, and Nigel Harris. 2008. Use of multiple wearable inertial sensors in upper limb motion tracking. Medical Engineering & Physics 30, 1 (Jan. 2008), 123–133. https://doi.org/10.1016/j.medengphy.2006.11. 010

    2008 doi

  55. [63]

    Xinyu Yi, Yuxiao Zhou, Vladislav Golyanik, Marc Habermann, Soshi Shimada, Christian Theobalt, and Feng Xu. 2022. Physical Inertial Poser (PIP): Physics- aware Real-time Human Motion Tracking from Sparse Inertial Sensors. (June 2022), 15

    2022

  56. [68]

    Yoon, Bong-Soo Kang, and Edward J

    Shaghayegh Zihajehzadeh, Paul K. Yoon, Bong-Soo Kang, and Edward J. Park

  57. [2015]

    IEEE Transactions on Instrumentation and Measurement 64, 12 (Dec

    UWB-Aided Inertial Motion Capture for Lower Body 3-D Dynamic Activity and Trajectory Tracking. IEEE Transactions on Instrumentation and Measurement 64, 12 (Dec. 2015), 3577–3587. https://doi.org/10.1109/TIM.2015.2459532 11

    2015

  58. [2017]

    http://arxiv.org/abs/1703.08014 arXiv:1703.08014 [cs]

    Sparse Inertial Poser: Automatic 3D Human Pose Estimation from Sparse IMUs. http://arxiv.org/abs/1703.08014 arXiv:1703.08014 [cs]

    arXiv

  59. [2018]

    A World of Wireless, Mobile and Multimedia Networks

    Enabling Runtime Adaptation of Physical Layer Settings for Dependable 9 UIST ’23, October 29-November 1, 2023, San Francisco, CA, USA DeVrio & Mollyn, et al. UWB Communications. In 2018 IEEE 19th International Symposium on "A World of Wireless, Mobile and Multimedia Networks" ...

    2023

  60. [2023]

    In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems (CHI ’23)

    IMUPoser: Full-Body Pose Estimation using IMUs in Phones, Watches, and Earbuds. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems (CHI ’23) . Association for Computing Machinery, New York, NY, USA, 1–12. https://doi.org/10.1145/3544548.3581392

    2023

  61. [2166]

    https://doi.org/10.1145/2470654.2481296

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.