Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

DoRF: Doppler Radiance Fields for Robust Human Activity Recognition Using Wi-Fi

T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read This paper claims that reconstructing a 3D velocity field from Wi-Fi Doppler projections and uniformly sampling it over a sphere — a Doppler radiance field (DoRF) — yields a view-invariant motion representation that improves cross-user…

desk verdict A modest, plausible extension of MORIC whose central view-invariance claim is undermined by a rotation gauge in the factorization, and whose reported gain is within noise. read the letter →

arxiv 2507.12132 v1 pith:KJ4ZRPSU submitted 2025-07-16 eess.SP cs.CV

classification eess.SPcs.CV
keywords Wi-FisensinghumanactivityrecognitionchannelstateinformationDopplervelocityneuralradiancefieldsfieldcross-usergeneralizationmultipath
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 tries to establish that Wi-Fi signals alone can support a view-invariant, three-dimensional description of human motion, and that this description improves cross-user activity recognition. The authors reconstruct a 3D velocity field from one-dimensional Doppler projections extracted from Wi-Fi channel state information, then re-sample that field onto a uniform grid of directions over a sphere, calling the result a Doppler radiance field (DoRF). On a four-gesture, six-person hand-motion dataset, they report leave-one-subject-out accuracy of 54.8% with standard deviation 7.2%, beating the MORIC baseline at 51.5% and all compared methods. The significance would be that Wi-Fi sensing could become more reliable for practical deployment across new users without retraining.

What carries the argument

The load-bearing object is the Doppler radiance field P(s, m, n) = v(s)^T d_mn, where v(s) is a reconstructed 3D velocity vector and d_mn are uniformly spaced unit directions on the sphere. The DoRF is obtained by solving a non-convex factorization Vr = V R^T + N via alternating least squares with unit-norm direction constraints, Procrustes alignment, and a DTW-based loss, then projecting each recovered velocity vector onto the full spherical grid. This converts an unordered, sparse set of one-dimensional Doppler observations into a dense, ordered, view-invariant tensor that is fed to the MORIC classifier, which applies random convolution kernels and max-pooling across the directional axis.

What would settle it

Place a motion-capture system beside the Wi-Fi setup and compare the recovered center-of-mass velocity v(s) with the ground-truth hand or body velocity during each gesture. If the factorization is correct, the two trajectories should align up to a fixed rotation; if the DoRF accuracy gain over MORIC disappears when the room's reflecting furniture is moved during a trial, that would indicate the time-invariant direction assumption is the source of the gain.

Watch

Extended reading notes

Core claim

The central claim is that a uniform Doppler radiance field, built by estimating the 3D center-of-mass velocity of motion from sparse Doppler projections and sampling it over a sphere, is a comprehensive and view-invariant motion representation. Rather than relying on whatever multipath reflection directions happen to be present, the method reconstructs a latent velocity trajectory V and a set of observation directions R from the observed projections Vr, then computes radial projections onto a fixed, dense spherical grid. Because the representation does not depend on the random, time-varying viewing geometry of the Wi-Fi access points, the authors argue it generalizes better to unseen users than representations built directly from raw Doppler projections. The reported 54.8% mean accuracy on UTHAMO, with lower variance than MORIC's 8.7%, is offered as evidence.

Load-bearing premise

The whole argument assumes that the multipath reflection directions stay fixed while the person moves, so that a single set of directions can explain all Doppler projections; in real rooms the directions shift with the body, and if that shift matters the reconstructed velocity field is distorted.

Editorial extensions

If this is right

  • Cross-user generalization of Wi-Fi human activity recognition improves, reducing the need for per-user calibration or retraining before deployment.
  • The same DoRF pipeline can fuse data from multiple antennas or access points by constructing a field per antenna and merging them before classification, making it robust to AP configuration.
  • Because DoRF encodes directional motion, it may support finer-grained tasks than classification, such as motion reconstruction or velocity estimation of the moving body.
  • The representation is designed to be insensitive to environmental multipath geometry, since the spherical sampling removes dependence on actual reflection directions.
  • The reported gain of 3.3 percentage points over MORIC, with a lower standard deviation (7.2% vs. 8.7%), suggests more stable performance across different users.

Reading between the lines

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

  • If the recovered velocity v(s) truly approximates the center-of-mass velocity, DoRF could serve as a bridge between Wi-Fi CSI and body-joint or hand-trajectory estimation, not just gesture classification.
  • The time-invariant direction assumption is the natural failure point: a natural extension is a time-varying or smoothly evolving set of directions R, which would test whether the remaining accuracy gap is caused by this approximation.
  • The uniform sampling grid size M acts as a resolution knob; one could test whether accuracy saturates as M increases and whether adaptive, information-driven grids perform better.
  • Applying DoRF to larger, multi-environment datasets with more subjects would clarify whether the 3.3-point gain holds beyond the single office setting and six users of UTHAMO.
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

3 major / 5 minor

Summary. The paper proposes Doppler Radiance Fields (DoRF), a Wi-Fi CSI-based motion representation for human activity recognition. The method first estimates a 3D velocity sequence V from 1D Doppler projections via alternating constrained least squares (Eq. (9)), then projects each velocity vector onto a uniform spherical grid to form P(s,m,n), and finally feeds this DoRF into the MORIC classifier. On the UTHAMO dataset with leave-one-subject-out evaluation, the method reports mean accuracy 54.8% (SD 7.2%) versus 51.5% (SD 8.7%) for MORIC and lower accuracy for other baselines. The paper claims that uniform sampling provides a view-invariant, complete motion representation that improves cross-user generalization.

Significance. If the claimed improvement were statistically reliable, the paper would contribute a novel way to turn sparse Doppler projections into a structured, dense representation for Wi-Fi HAR, and the LOSO evaluation on a public dataset is a strength. The algorithmic description is mostly complete, and comparison against several baselines is useful. However, the evidence for the central claim is currently weak: the accuracy gap is within the reported standard deviations, no significance testing or ablation is provided, and the factorization underlying the DoRF has an unresolved rotational gauge freedom that undermines the well-definedness of the proposed representation. The idea is promising but requires substantial additional validation.

major comments (3)
  1. [Section 3.2, Eq. (9) and Algorithm 1] The optimization objective and both regularizers are invariant to the simultaneous transformation (V,R) -> (VQ,RQ) for any orthogonal Q, since V_r = V R^T is unchanged, ||v(s)|| is unchanged, and ||r_i||=1 is preserved. The Procrustes step (Algorithm 1, line 5) is described only as 'stabilize orientation' and, without an external reference frame, cannot remove this gauge freedom. Under the DoRF construction of Section 3.3, P(s,m,n)=v(s)^T d_{mn} becomes v(s)^T Q d_{mn}, which for a general Q is not a permutation of the fixed grid values. Therefore the claimed view-invariant, complete representation is not guaranteed and different random initializations of R can produce different DoRFs and different classification inputs. The authors should either fix a canonical orientation (e.g., Procrustes against a fixed template or SVD-based alignment of V) or use a rotation-invariant feature representation, and report sensitivity of accuracy to initialization.
  2. [Section 4.3, Table 1] The reported improvement of 54.8% over 51.5% is less than the subject-level standard deviations (7.2% and 8.7%), and no significance test, confidence interval, or per-user results are reported. With only six users in LOSO, this difference could be within chance fluctuation. The claim that uniform sampling 'noticeably enhances' generalization is therefore not supported by the evidence. The authors should report paired per-user accuracy, a statistical test (e.g., Wilcoxon signed-rank or paired bootstrap), and effect sizes. They should also include an ablation that separates the contribution of the 3D velocity reconstruction (Section 3.2) from the uniform sphere sampling (Section 3.3) by comparing, e.g., raw MORIC projections, reconstructed velocity sequence used directly, and the full DoRF.
  3. [Section 3.2, Eqs. (5)-(9)] The model assumes a single time-invariant set of unit directions R such that V_r(s,i)=v(s)^T r_i + n(s,i). In a multipath environment, the effective reflection directions move with the body and change over time; the factorization is then not the physical model implied by Eq. (5). This matters because the DoRF is built from the recovered v(s); if the model is misspecified, the 'velocity' is a latent factor rather than an actual motion estimate. The authors should either justify the approximation with measurements (e.g., reconstruction error or comparison with an independent motion capture), or soften the physical claims and treat the factorization as a learned low-rank representation. A simple diagnostic would be to report the fit of Eq. (9) on held-out data and the sensitivity of accuracy to the number of retained directions N.
minor comments (5)
  1. [Section 3.2, Algorithm 1, line 6] The loss is described as 'DTW-based', but Eq. (9) uses squared error and no DTW definition is given; please specify the exact loss and how DTW is incorporated into the alternating updates.
  2. [Section 4.2] The regularization parameters λ and γ in Eq. (9) are not reported; without them the experiments are not reproducible.
  3. [Section 4.2/4.3] Please clarify how the baselines in Table 1 were obtained—whether they are re-implemented with the same preprocessing and training setup or copied from the original papers—and report per-user accuracies.
  4. [References [14] and [18]] References [14] and [18] are arXiv preprints; if the methods/dataset have been published elsewhere, please cite the final versions.
  5. [Figure 2] Figure 2 is not referenced in the text at its first occurrence; please add a cross-reference and define the color coding.

Circularity Check

0 steps flagged · score 2.0 of 10

No load-bearing circularity; the headline LOSO accuracy is an independent empirical comparison, with only minor reliance on authors' prior work.

full rationale

The paper's derivation chain is: CSI -> delay-Doppler projections Vr (Eq. 8) -> matrix factorization Vr = V R^T + N (Eq. 9) -> uniform spherical sampling P(s,m,n)=v(s)^T d_mn (Eq. 16) -> MORIC classifier. Each stage is either a defined transformation or a fitted latent factor, not a prediction of a quantity that is itself an input. The V,R factorization is estimated from the observed Vr by alternating least squares with regularizers; this is an unsupervised reconstruction/denoising step. The DoRF is then a deterministic function of the fitted V, and the reported 54.8% LOSO accuracy is evaluated on held-out subjects against MORIC and other baselines, so the central claim does not reduce to the method's construction. The main self-references (MORIC [14] for the classifier and UTHAMO [18] for the data) are used as a baseline and as a dataset, not as the justification of the claimed improvement; comparing against the authors' own baseline is still an external empirical test. The strongest challenge to the method—the rotational gauge freedom in Eq. (9), where the objective is invariant to (V,R)->(VQ,RQ) while Algorithm 1's Procrustes step has no external reference frame—is an identifiability/well-posedness concern, not a circularity: it does not make any output equal to its input by construction. The paper is therefore self-contained with respect to its experimental claim; the score reflects only minor non-load-bearing self-citation.

Assumptions & free parameters 3 free parameters · 4 assumptions · 1 invented entities

The central model rests on a low-rank factorization whose identifiability depends on time-invariant observation directions and sufficient directional coverage. Regularization parameters lambda and gamma are introduced but never specified. The DoRF itself is a derived representation, not a new physical entity.

free parameters (3)
  • lambda
    Regularization strength for velocity magnitude in Eq. (9); value not specified in the paper.
  • gamma
    Regularization strength for direction updates in Eq. (9); value not specified in the paper.
  • M = 8
    Grid size for uniform sampling of the sphere (Section 4.2); a hyperparameter chosen by hand.
assumptions (4)
  • domain assumption Each Doppler projection equals the inner product of a common 3D velocity vector and a fixed unit direction (Eqs. 8-9).
    Assumes the direction r_i of each multipath is constant over the time window, which is not stated in the paper and is likely false in a dynamic indoor environment. Invoked in Section 3.2.
  • domain assumption The set of delay bins provides at least three non-coplanar directions so that the velocity is identifiable.
    The factorization can only recover v(s) if the directions span R3; the paper does not verify this on real data. This is the geometric identifiability assumption behind Algorithm 1.
  • standard math The IDFT of CSI yields delay bins that correspond to distinct multipath components with resolvable Doppler peaks.
    Equations (3)-(8) rely on standard Fourier analysis, but real CSI has limited bandwidth and noise, so bins are not perfectly separated.
  • ad hoc to paper A single global rotation (Procrustes) can align the recovered motion basis across trials.
    Section 3.2 and Algorithm 1 step 5 assume each trial's recovered directions are related by a rotation, which is not justified for arbitrary gestures.
invented entities (1)
  • Doppler radiance field (DoRF)
    purpose: To represent motion as dense directional velocity projections on a uniform sphere grid for classification.
    The DoRF is a deterministic projection of the estimated velocity field, not a physically new entity; it has no falsifiable predictions beyond the activity classification performance.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DoRF: Doppler Radiance Fields for Robust Human Activity Recognition Using Wi-Fi." pith.science (2026). https://pith.science/paper/KJ4ZRPSU

@misc{pith2026250712132,
  author       = {Pith},
  title        = {Pith review of: DoRF: Doppler Radiance Fields for Robust Human Activity Recognition Using Wi-Fi},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KJ4ZRPSU}},
  note         = {Machine review of arXiv:2507.12132}
}
read the original abstract

Wi-Fi Channel State Information (CSI) has gained increasing interest for remote sensing applications. Recent studies show that Doppler velocity projections extracted from CSI can enable human activity recognition (HAR) that is robust to environmental changes and generalizes to new users. However, despite these advances, generalizability still remains insufficient for practical deployment. Inspired by neural radiance fields (NeRF), which learn a volumetric representation of a 3D scene from 2D images, this work proposes a novel approach to reconstruct an informative 3D latent motion representation from one-dimensional Doppler velocity projections extracted from Wi-Fi CSI. The resulting latent representation is then used to construct a uniform Doppler radiance field (DoRF) of the motion, providing a comprehensive view of the performed activity and improving the robustness to environmental variability. The results show that the proposed approach noticeably enhances the generalization accuracy of Wi-Fi-based HAR, highlighting the strong potential of DoRFs for practical sensing applications.

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. WiFuse: An Attention Mechanism for Human Activity Recognition using Fused CSI Amplitude and Delay-Doppler Channel Features

    cs.CV 2026-08 conditional novelty 4.0 of 10

    A dual-stream CSI framework fusing amplitude and Delay-Doppler features with a ResNet-TCN attention model reaches 95.28% on XRF55 and 98.20% on Wi-MIR.

Reference graph

Works this paper leans on

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

  1. [1]

    Using existing Wi-Fi infrastructure eliminates the need for additional hardware, making deployment more accessible and cost-effective

    INTRODUCTION Wi-Fi sensing has emerged as a practical approach for indoor human activity recognition and tracking, offering distinct advan- tages over traditional methods such as cameras and wearable de- vices [1]. Using existing Wi-Fi infrastructure eliminates the need for additional hardware, making deployment more accessible and cost-effective. Unlike ...

  2. [2]

    DELAY-DOPPLER DECOMPOSITION OF CSI Consider a narrow-band Wi-Fi link with N OFDM subcarriers of spacing ∆f and carrier frequency fc. At time s, the channel–state matrix at subcarrier n is modeled as Hn(s) = L−1X l=0 βl(s) e−j2πfn τl(s), (1) fn = fc − n − N 2 ∆f, (2) where Hn(s) is the complex CSI at subcarrier n and time s; βl(s) and τl(s) are the complex...

  3. [3]

    METHOD This section describes CSI phase pre-processing, the modeling framework for 3D velocity estimation of the motion center of mass, the transformation into a uniform DoRF, and the activity classifica- tion method used in this work. Fig. 1 shows the different steps of the proposed method. 3.1. CSI Phase Sanitization CSI phase measurements are distorted...

  4. [4]

    Data The UTHAMO dataset, introduced in [18], is utilized to assess the performance of the proposed method

    EXPERIMENT 4.1. Data The UTHAMO dataset, introduced in [18], is utilized to assess the performance of the proposed method. CSI data was collected from six participants performing four gestures— circle, left–right, up–down, and push–pull—using an ASUS RT-AC 86U router equipped with three antennas in a static indoor office environ- ment of size 6 m × 5.6 m....

  5. [5]

    CONCLUSION This study introduces a novel Wi-Fi-based HAR method that con- structs a unique motion representation by uniformly sampling the radiance fields extracted from CSI. Inspired by NeRF, it recon- structs a 3D latent motion representation from one-dimensional Doppler velocity projections, capturing dynamic motion while sup- pressing static environme...

  6. [6]

    A tutorial-cum- survey on self-supervised learning for wi-fi sensing: Trends, challenges, and outlook,

    Ahmed Y Radwan, Mustafa Yildirim, Navid Hasanzadeh, Hina Tabassum, and Shahrokh Valaee, “A tutorial-cum- survey on self-supervised learning for wi-fi sensing: Trends, challenges, and outlook,” IEEE Communications Surveys & Tutorials, 2025

  7. [7]

    Litehar: Lightweight human activity recognition from wifi signals with random convolution kernels,

    Hojjat Salehinejad and Shahrokh Valaee, “Litehar: Lightweight human activity recognition from wifi signals with random convolution kernels,” inICASSP 2022-2022 IEEE In- ternational Conference on Acoustics, Speech and Signal Pro- cessing (ICASSP). IEEE, 2022, pp. 4068–4072

  8. [8]

    A multiple wifi sensors assisted human activity recognition scheme for smart home,

    Jianyang Ding, Yong Wang, Qian Xie, and Jiajun Niu, “A multiple wifi sensors assisted human activity recognition scheme for smart home,” IEEE Sensors Journal, 2024

Show all 28 references
  1. [9]

    A survey on behavior recognition using wifi channel state information,

    Siamak Yousefi, Hirokazu Narui, Sankalp Dayal, Stefano Er- mon, and Shahrokh Valaee, “A survey on behavior recognition using wifi channel state information,” IEEE Communications Magazine, vol. 55, no. 10, pp. 98–104, 2017

  2. [10]

    Fresnel zone-based voting with capsule networks for human activity recognition from channel state information,

    Radomir Djogo, Hojjat Salehinejad, Navid Hasanzadeh, and Shahrokh Valaee, “Fresnel zone-based voting with capsule networks for human activity recognition from channel state information,” IEEE Internet of Things Journal, pp. 1–1, 2024

  3. [11]

    A study on the lightweight and fast response gru techniques for indoor continuous motion recognition based on wi-fi csi,

    Kyongseok Jang, Chao Sun, Junhao Zhou, Yongbin Seo, Youngok Kim, and Seyeong Choi, “A study on the lightweight and fast response gru techniques for indoor continuous motion recognition based on wi-fi csi,” IEEE Access, 2025

  4. [12]

    Csi-glstn: A location- independent csi human activity recognition method based on spatio-temporal and channel feature fusion,

    Yong Zhang, Dingchao Wu, Yujie Wang, Yang Zhang, Guangkun Jiand, and Jiaqiu Ai, “Csi-glstn: A location- independent csi human activity recognition method based on spatio-temporal and channel feature fusion,” IEEE Transac- tions on Instrumentation and Measurement, 2024

  5. [13]

    Rosefi: A robust sedentary behavior monitoring system with commodity wifi devices,

    Cheng Peng, Linqing Gui, Biyun Sheng, Zhengxin Guo, and Fu Xiao, “Rosefi: A robust sedentary behavior monitoring system with commodity wifi devices,” IEEE Transactions on Mobile Computing, vol. 23, no. 5, pp. 6470–6489, 2023

  6. [14]

    Wifi csi-based device-free sensing: from fresnel zone model to csi-ratio model,

    Dan Wu, Youwei Zeng, Fusang Zhang, and Daqing Zhang, “Wifi csi-based device-free sensing: from fresnel zone model to csi-ratio model,” CCF Transactions on Pervasive Comput- ing and Interaction, pp. 1–15, 2022

  7. [15]

    Fewsense, towards a scalable and cross-domain wi-fi sensing system using few-shot learning,

    Guolin Yin, Junqing Zhang, Guanxiong Shen, and Yingying Chen, “Fewsense, towards a scalable and cross-domain wi-fi sensing system using few-shot learning,” IEEE Transactions on Mobile Computing, vol. 23, no. 1, pp. 453–468, 2022

  8. [16]

    Exposing data leakage in wi-fi csi-based human action recognition: A critical analysis,

    Domonkos Varga, “Exposing data leakage in wi-fi csi-based human action recognition: A critical analysis,” Inventions, vol. 9, no. 4, pp. 90, 2024

  9. [17]

    Sharp: Environment and person independent activity recognition with commod- ity ieee 802.11 access points,

    Francesca Meneghello, Domenico Garlisi, Nicol`o Dal Fabbro, Ilenia Tinnirello, and Michele Rossi, “Sharp: Environment and person independent activity recognition with commod- ity ieee 802.11 access points,” IEEE Transactions on Mobile Computing, vol. 22, no. 10, pp. 6160–6175, 2022

  10. [18]

    Widar3. 0: Zero-effort cross-domain gesture recognition with wi-fi,

    Yi Zhang, Yue Zheng, Kun Qian, Guidong Zhang, Yunhao Liu, Chenshu Wu, and Zheng Yang, “Widar3. 0: Zero-effort cross-domain gesture recognition with wi-fi,” IEEE Transac- tions on Pattern Analysis and Machine Intelligence , vol. 44, no. 11, pp. 8671–8688, 2021

  11. [19]

    Moric: Csi delay- doppler decomposition for robust wi-fi-based human activity recognition,

    Navid Hasanzadeh and Shahrokh Valaee, “Moric: Csi delay- doppler decomposition for robust wi-fi-based human activity recognition,” arXiv preprint arXiv:2506.12997, 2025

  12. [20]

    Nerf: Representing scenes as neural radiance fields for view synthe- sis,

    Ben Mildenhall, Pratul P Srinivasan, Matthew Tancik, Jonathan T Barron, Ravi Ramamoorthi, and Ren Ng, “Nerf: Representing scenes as neural radiance fields for view synthe- sis,” Communications of the ACM, vol. 65, no. 1, pp. 99–106, 2021

  13. [21]

    Decimeter ranging with chan- nel state information,

    Navid Tadayon, Muhammed Tahsin Rahman, Shuo Han, Shahrokh Valaee, and Wei Yu, “Decimeter ranging with chan- nel state information,” IEEE Transactions on Wireless Com- munications, vol. 18, no. 7, pp. 3453–3468, 2019

  14. [22]

    Procrustes analysis,

    Amy Ross, “Procrustes analysis,” Course report, Department of Computer Science and Engineering, University of South Carolina, vol. 26, pp. 1–8, 2004

  15. [23]

    Uthamo: A multi-modal wi-fi csi-based hand motion dataset,

    Navid Hasanzadeh, Radomir Djogo, Hojjat Salehinejad, and Shahrokh Valaee, “Uthamo: A multi-modal wi-fi csi-based hand motion dataset,” 2025

  16. [24]

    Free your csi: A channel state information extraction platform for modern wi-fi chipsets,

    Francesco Gringoli, Matthias Schulz, Jakob Link, and Matthias Hollick, “Free your csi: A channel state information extraction platform for modern wi-fi chipsets,” in Proceed- ings of the 13th International Workshop on Wireless Network Testbeds, Experimental Evaluation & Charac...

  17. [25]

    Decoupled weight decay regularization,

    Ilya Loshchilov and Frank Hutter, “Decoupled weight decay regularization,” arXiv preprint arXiv:1711.05101, 2017

  18. [26]

    Melon: Nerf with unposed images in so (3),

    Axel Levy, Mark Matthews, Matan Sela, Gordon Wetzstein, and Dmitry Lagun, “Melon: Nerf with unposed images in so (3),” arXiv preprint arXiv:2303.08096, 2023

  19. [27]

    Joint human orientation-activity recogni- tion using wifi signals for human-machine interaction,

    Hojjat Salehinejad, Navid Hasanzadeh, Radomir Djogo, and Shahrokh Valaee, “Joint human orientation-activity recogni- tion using wifi signals for human-machine interaction,” in ICASSP 2023-2023 IEEE international conference on acous- tics, speech and signal processing (ICASSP)....

  20. [28]

    Enhancing gen- eralization in human activity recognition through improved wi-fi channel state information phase processing and antenna pair selection,

    Navid Hasanzadeh and Shahrokh Valaee, “Enhancing gen- eralization in human activity recognition through improved wi-fi channel state information phase processing and antenna pair selection,” in 2024 IEEE 34rd International Workshop on Machine Learning for Signal Processing (ML...

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.