WiFi based Human Fall and Activity Recognition using Transformer based Encoder Decoder and Graph Neural Networks

Andrea Merello; Arash Ajoudani; Elisa Motta; Marco Crepaldi; Marta Lagomarsino; Olivia Nocentini; Younggeol Cho

REVIEW 4 major objections 6 minor 21 references

WiFi based Human Fall and Activity Recognition using Transformer based Encoder Decoder and Graph Neural Networks

T0 review · 4 major / 6 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read WiFi signals alone can reconstruct a human skeleton and detect falls with accuracy close to camera-based systems.

desk verdict A useful WiFi-skeleton pipeline and a new fall dataset, but the headline fall-detection accuracy is not tested across users because the split is within-subject. read the letter →

arxiv 2504.16655 v1 pith:TOPVGQW7 submitted 2025-04-23 cs.CV

Younggeol Cho , Elisa Motta , Olivia Nocentini , Marta Lagomarsino , Andrea Merello , Marco Crepaldi , Arash Ajoudani This is my paper · ORCID

classification cs.CV

keywords humanskeletonposeestimationactionrecognitionchannelstateinformation(CSI)transformernetworkgraphneuralnetworksfalldetectionWi-Fisensingprivacy-preservingmonitoring

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 sets out to show that Wi-Fi Channel State Information (CSI) — the signal data already available from commodity Wi-Fi hardware — carries enough information to reconstruct a person's 2D skeleton, and that those skeletons can drive activity recognition and fall detection. It introduces TED-Net, a Transformer-based encoder-decoder that maps CSI tensors to 17 body-keypoint coordinates, and couples it with a Directed Graph Neural Network that classifies stand, walk, squat, and fall at the frame level. On the public multi-modal benchmark, TED-Net reports a PCK10 of 73.8%, above the 63.5% reported for the prior CSI pose estimator. On the authors' own fall dataset, CSI-derived skeletons give 90.5% fall accuracy, compared with 94.3% when the same graph network is fed RGB-derived skeletons. If these results hold, home Wi-Fi alone could provide privacy-preserving fall monitoring without cameras.

What carries the argument

The load-bearing mechanism is TED-Net's CNN-Transformer encoder-decoder: three convolutional encoders per receiving antenna compress the CSI tensor, the encoded features from three antennas are concatenated and reshaped into a sequence of 34 tokens (17 keypoints × 2 coordinates), and a two-layer, eight-head Transformer applies self-attention across that sequence so globally distant joints can influence each other; a convolutional decoder then up-samples the attended features and outputs the 17 (x, y) coordinates. The second mechanism is the customized DGNN, which uses directed edges between keypoints, a dynamically adjusted adjacency matrix, and temporal convolution to classify each frame as stand, walk, squat, or fall. The Transformer's global cross-joint attention is what lets the network infer coordinated body poses from spatially coarse CSI, and the DGNN's directed graph structure is what lets it propagate joint information along anatomical connections to separate falls from similar postures such as squats.

What would settle it

Retrain TED-Net and the prior baseline on the full public dataset with the same split and PCK normalization; if TED-Net's PCK10 no longer exceeds the baseline's, the superiority claim fails.

Watch

Extended reading notes

Core claim

The central claim is that a Transformer-based encoder-decoder can estimate 17 two-dimensional skeleton keypoints directly from raw CSI and that those keypoints are accurate enough for downstream action recognition. The paper reports that TED-Net outperforms prior CSI-based pose estimators on the public dataset, with an average PCK10 of 73.8% versus 63.5%, and that the resulting skeletons preserve body-center joints best (pelvis 99.3%, shoulders about 85%) while hands are hardest. On the custom fall dataset, the DGNN achieves 90.5% accuracy on falls from CSI skeletons and 94.3% from RGB skeletons, and the paper argues this gap is small enough that CSI skeletons can replace camera-based skeletons for fall detection. The paper also claims that pose estimation is not subject-dependent and that the system can track pelvis/torso position with under 10% normalized error.

Load-bearing premise

The comparison to prior work assumes that the published PCK numbers were produced under the same evaluation protocol — same subject subset, same train/test split, same torso-length normalization — as the paper's 10-subject subset.

Editorial extensions

If this is right

A home Wi-Fi setup with one transmitter and three receivers could run continuous fall detection without cameras, preserving privacy in bathrooms and bedrooms.
CSI-derived skeletons can replace RGB-derived skeletons for frame-level action recognition in semi-constrained indoor environments at a small accuracy cost.
The DGNN's confusion matrix shows fall misclassifications mostly as squat, so adding a dedicated squat class keeps false alarms low.
Torso and pelvis tracking with under 10% normalized error enables coarse position monitoring alongside action recognition.
The architecture can be extended to other healthcare monitoring tasks that rely on large-scale body dynamics rather than fine joint detail.

Reading between the lines

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

Beyond the paper: the training loop still requires RGB-skeleton supervision from a camera, so the system is privacy-preserving at inference time only; deployment would need a one-time camera-based calibration phase.
Beyond the paper: the reported 90.5% fall accuracy is measured on the authors' own data collected in a fixed room with a fixed transmitter-receiver geometry; moving furniture, another person, or a different room layout would likely degrade performance, so the claim is for semi-constrained settings only.
Beyond the paper: because pelvis tracking error is the lowest, a simpler detector could use only torso/pelvis trajectory to flag falls, potentially removing the need for full-keypoint graph classification.
Beyond the paper: a natural stress test is to evaluate TED-Net on the full public dataset with all subjects and across multiple environments rather than the 10-subject single-environment subset; if the gap over prior work narrows, the claimed advantage may be protocol-dependent.

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.

Desk Editor's Note

read the letter

Short version: this paper builds a plausible CSI-to-skeleton estimator (TED-Net) and a DGNN action classifier, and contributes a new 20-subject fall dataset. The pose-estimation results on MM-Fi are decent, and the paper is honest that fall-time skeletons are much worse than non-fall ones (PCK10 28.5% vs 83.4%). What it does not do is demonstrate cross-user fall detection: all splits are within-subject, so every test subject is in training for both TED-Net and the DGNN. That is the load-bearing issue.

The architecture is a reasonable combination of known components: convolutional encoders, a two-layer transformer, a decoder, and a directed GNN with temporal windows. No claim of fundamental novelty, and that is fine. The dataset is the real contribution: synchronized CSI from three antennas plus RGB at 30 Hz for 20 subjects doing stand, walk, squat, and two kinds of falls, with a careful calibration check of the CSI chain. That is worth having.

The main soft spot is the evaluation protocol. For the custom dataset, standing/walking/squatting are split by duration 80/20 per subject, and fall trials are split by repetition (first four train, fifth test) per subject. So the 90.5% CSI fall accuracy and the 94.8% average are within-subject numbers. The model can memorize subject-specific CSI-to-skeleton mappings and pose dynamics; an unseen-user fall might behave far worse. The text even says 'TED-Net is not subject-dependent' based on two example subjects that are both in training. That claim does not follow.

Second soft spot: the comparison with MetaFi++ [13] and MetaFi [16] in Table III is not controlled. The authors use a 10-subject subset of MM-Fi with their own 80/20 split, while the cited numbers come from those papers' protocols. Unless the normalization and subset are matched, the 73.8 vs 63.5 PCK10 margin is not interpretable. The reader flagged this and I agree.

Third, the ground truth for PCK is YOLOv11Pose output, not an independent measurement. That is a minor point for comparing methods under the same GT, but it means the absolute PCK values should be taken as 'agreement with a model,' not accuracy against a marker-based system.

The paper is clearly written, and the discussion of fall-time pose degradation is refreshingly honest. The central engineering question—whether WiFi-only skeletons can drive fall detection—remains plausible but untested across people. A serious referee would want a cross-subject split, matched baselines, and ideally code/data release. I would not desk-reject it; the dataset and pipeline justify a round of review, but the current claims are overstated.

Referee Report

4 major / 6 minor

Summary. The paper proposes TED-Net, a CNN-transformer encoder-decoder that estimates 17 2D skeleton keypoints from Wi-Fi CSI tensors captured by three antennas, and a customized Directed Graph Neural Network (DGNN) that classifies the resulting skeleton sequences into stand, walk, squat, and fall. TED-Net is trained with keypoints extracted from synchronized RGB frames by YOLOv11Pose as supervision. The method is evaluated on a 10-subject subset of the public MM-Fi dataset and on a newly collected 20-subject dataset that includes fall and non-fall actions. The paper reports an average PCK10 of 73.8% on MM-Fi, compares this with published values from MetaFi++ and MetaFi, and reports CSI-skeleton action-recognition accuracy of 90.5% for fall detection versus 94.3% for RGB-skeleton input.

Significance. If the headline results were obtained under matched protocols and subject-disjoint evaluation, the paper would offer a practically relevant contribution to privacy-preserving fall detection: the architecture is clearly described, the integrated pose-estimation-plus-action-recognition pipeline is sensible, and the custom hardware setup is reported in enough detail to be replicated. The explicit PCK tables and the inclusion of both fall and fall-like actions (squat) are useful. However, the current evidence does not establish the central claims, because the comparison with prior work is not controlled and the fall-detection evaluation is performed on a within-subject split in which every test subject also appears in training. These issues are fixable within the paper's scope, but they require additional experiments and a more cautious interpretation of the results.

major comments (4)

[Section IV-A, Table III] The comparison with MetaFi++ [13] and MetaFi [16] is not controlled. TED-Net is evaluated on a 10-subject subset of MM-Fi in one environment with an 80/20 temporal split described in Section III-B, while the cited PCK values are taken from the original papers, which used their own subject subsets, environments, and evaluation protocols. Because PCK depends on the subject pool, action set, and normalization, the reported margin at PCK10 (73.8% vs. 63.5%) may be a protocol artifact rather than an architectural advantage. The claim that TED-Net outperforms existing approaches therefore needs either a same-protocol re-implementation of the baselines or an explicit statement that the comparison is only indicative.
[Section III-B, Table V, Figure 4] The fall-detection evaluation uses a within-subject split: for each subject, the first four fall repetitions are used for training and the fifth repetition for testing, so every test subject is also present in the training set of both TED-Net and the DGNN. The reported CSI fall accuracy of 90.5% and average accuracy of 94.8% do not demonstrate that a fall by an unseen user would be detected. To support the claimed applicability to home fall detection, the authors should report leave-one-subject-out or otherwise subject-disjoint results, per-subject accuracy, and confidence intervals or variance over repeated runs.
[Section IV-A, Table III] The text states that TED-Net is 'superior across all PCK metrics' compared with [16] and outperforms [13], but the table itself shows that TED-Net is better than [13] only at PCK10 (73.8% vs. 63.5%), nearly identical at PCK20 (87.0% vs. 86.7%), and worse at PCK30, PCK40, and PCK50 (91.9/94.5/96.0 vs. 93.1/95.9/97.3). The claim as written is contradicted by the paper's own table and should be revised, or the higher-threshold behavior should be explained.
[Section III-A and IV-A] The skeleton keypoints labeled as ground truth are produced by YOLOv11Pose from RGB images, so the PCK scores measure agreement with a learned pose estimator rather than with an independent physical measurement of joint positions. Missing or displaced keypoints are also corrected by interpolation, which can bias the evaluation. The paper should explicitly acknowledge this limitation and, ideally, validate a subset of the keypoints against manually annotated or marker-based data.

minor comments (6)

[Section III-B] The sentence 'an average age of ±29 years and an average height of ±172 cm' appears to be missing the means and standard deviations; please report the actual values, e.g., mean ± std.
[Section III-C1] The definition of PCK is slightly inaccurate: the text says α is 'the minimum required proportion of correctly detected joints within a skeleton,' but Equation (1) defines PCK as the fraction of keypoints whose normalized distance is below α. Please align the wording with the equation.
[Section III-A and Table I] The input shape is described as (10, 3, 114) in the text but as (1, 114, 10) per receiver in Table I, and the receiver axis is not clearly indicated. Please clarify the tensor layout and any permutation or reshaping between acquisition and network input.
[Section IV-A, Figure 3] The statement that two example subjects demonstrate that TED-Net 'is not subject-dependent' is an overgeneralization from two individuals; please qualify this claim or provide quantitative per-subject results.
[Table IV] The meaning of the X and O columns is explained only in the text; please define these symbols directly in the table caption for clarity.
[Contributions, Section I] The phrase 'preliminary artificial dataset' is confusing; the dataset is a real collected dataset, so 'custom' or 'in-house' would be more accurate.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper is an empirical supervised-learning pipeline whose claims are tested on held-out temporal splits and against external prior work, not reduced to its own inputs by construction.

full rationale

The paper contains no derivation step in which an output is defined in terms of its own prediction, no fitted parameter is renamed as a prediction, and no load-bearing argument rests on a self-citation chain. The TED-Net regression target (YOLOv11Pose keypoints from RGB images) is an external teacher model, not the CSI model's own output, so PCK measures agreement with that teacher on held-out frames rather than being forced by construction. The comparison against MetaFi++ [13] and MetaFi [16] assumes that the cited PCK values were computed under a matching protocol (same MM-Fi subset, split, and normalization); if that assumption fails, the comparison is weakened, but that is a validity/protocol concern, not circularity. The custom-dataset evaluation uses within-subject temporal splits (80/20 by duration; first four fall repetitions train and fifth tests), which may overestimate cross-user generalization for fall detection, but the evaluation is still a genuine empirical test with unseen time windows; the reported fall accuracy is not an algebraic consequence of the training labels. The DGNN comparison between CSI-derived and RGB-derived skeletons is a direct contrast and would fail if the CSI skeleton carried no discriminative information. No cited uniqueness theorem or prior-work premise is invoked to forbid alternative architectural choices, and the authors' own prior work is not used as the justification for the central claim. Hence no circular step can be exhibited from the paper's equations or citation structure.

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

The central claims rest on assumptions about the information content of CSI, the validity of YOLO-generated ground truth, the comparability of literature baselines, and the representativeness of within-subject evaluation. No new physical entities are introduced; the free parameters are architecture choices and the unspecified DGNN temporal window.

free parameters (5)

CSI input buffer window = 10 samples (30 Hz effective)
Chosen to match camera frame rate; determines temporal context and directly affects pose estimation input.
Transformer embedding dimension = 384
Chosen by hand; not justified or ablated.
Transformer layers and heads = 2 layers, 8 heads
Architecture choice; no ablation reported.
DGNN temporal window = not specified
The number of previous frames fed for frame-level action recognition is never stated, yet it is central to the DGNN design.
CNN kernel sizes and strides = e.g., kernel=(4,3), stride=(2,2) in encoder 1
Hand-designed; no systematic search.

assumptions (4)

domain assumption CSI amplitude variations encode sufficient information to reconstruct 2D human skeletons.
Inherited from the WiFi sensing literature and the premise of the entire paper; not derived here.
domain assumption YOLOv11Pose keypoints are an accurate ground truth for human pose.
The authors use YOLO output as ground truth for training and evaluation; any YOLO bias propagates into PCK and action labels.
domain assumption Reported PCK values from [13] and [16] come from the same evaluation protocol as the authors'.
Table III compares against literature numbers without demonstrating identical subsets, splits, or PCK normalization.
domain assumption Within-subject temporal splits are sufficient to support generalization claims for fall detection.
The custom dataset is split by duration and repetitions within the same subjects and environment, not by held-out participants.

how reviews work

0 comments

Cite this review

Pith. "Pith review of WiFi based Human Fall and Activity Recognition using Transformer based Encoder Decoder and Graph Neural Networks." pith.science (2026). https://pith.science/paper/TOPVGQW7

@misc{pith2026250416655,
  author       = {Pith},
  title        = {Pith review of: WiFi based Human Fall and Activity Recognition using Transformer based Encoder Decoder and Graph Neural Networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TOPVGQW7}},
  note         = {Machine review of arXiv:2504.16655}
}

read the original abstract

Human pose estimation and action recognition have received attention due to their critical roles in healthcare monitoring, rehabilitation, and assistive technologies. In this study, we proposed a novel architecture named Transformer based Encoder Decoder Network (TED Net) designed for estimating human skeleton poses from WiFi Channel State Information (CSI). TED Net integrates convolutional encoders with transformer based attention mechanisms to capture spatiotemporal features from CSI signals. The estimated skeleton poses were used as input to a customized Directed Graph Neural Network (DGNN) for action recognition. We validated our model on two datasets: a publicly available multi modal dataset for assessing general pose estimation, and a newly collected dataset focused on fall related scenarios involving 20 participants. Experimental results demonstrated that TED Net outperformed existing approaches in pose estimation, and that the DGNN achieves reliable action classification using CSI based skeletons, with performance comparable to RGB based systems. Notably, TED Net maintains robust performance across both fall and non fall cases. These findings highlight the potential of CSI driven human skeleton estimation for effective action recognition, particularly in home environments such as elderly fall detection. In such settings, WiFi signals are often readily available, offering a privacy preserving alternative to vision based methods, which may raise concerns about continuous camera monitoring.

Figures

Figures reproduced from arXiv: 2504.16655 by the authors.

**Figure 1.** Overview of the CSI-driven Human Skeleton Pose Estimation and Action Recognition. The network consists of two main parts: (Part A) CSI-based [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗

**Figure 2.** Experimental setup. The CSI signal was emitted from one transmitter [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** CSI-driven Human skeleton pose estimation result based on Ted-Net. This example illustrates two subjects performing the five actions, demonstrating [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: Confusion matrices showing recognition results based on (a) RGB [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

**Figure 5.** Figure 5: Normalized Tracking Error Across Body Segments. The position [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

21 extracted references · 18 canonical work pages

[13]

Metafi++: Wifi-enabled transformer-based human pose estimation for metaverse avatar simulation,

Y . Zhou, H. Huang, S. Yuan, H. Zou, L. Xie, and J. Yang, “Metafi++: Wifi-enabled transformer-based human pose estimation for metaverse avatar simulation,” IEEE Internet of Things Journal , vol. 10, no. 16, pp. 14 128–14 136, 2023

work page 2023
[16]

Metafi: Device-free pose estimation via commodity wifi for metaverse avatar simulation,

J. Yang, Y . Zhou, H. Huang, H. Zou, and L. Xie, “Metafi: Device-free pose estimation via commodity wifi for metaverse avatar simulation,” in 2022 IEEE 8th World F orum on Internet of Things (WF-IoT) . IEEE, 2022, pp. 1–6

work page 2022
[1]

Ultralytics yolo11,

G. Jocher and J. Qiu, “Ultralytics yolo11,” 2024. [Online]. Available: https://github.com/ultralytics/ultralytics 8

work page 2024
[2]

Openpose: Realtime multi-person 2d pose estimation using part affinity fields,

Z. Cao, G. Hidalgo, T. Simon, S.-E. Wei, and Y . Sheikh, “Openpose: Realtime multi-person 2d pose estimation using part affinity fields,” IEEE transactions on pattern analysis and machine intelligence , vol. 43, no. 1, pp. 172–186, 2019

work page 2019
[3]

Blazepose: On-device real-time body pose tracking,

V . Bazarevsky, I. Grishchenko, K. Raveendran, T. Zhu, F. Zhang, and M. Grundmann, “Blazepose: On-device real-time body pose tracking,” arXiv preprint arXiv:2006.10204 , 2020

arXiv 2006
[4]

Real move: Effortless and precise human movement in real- time,

R. Move, “Real move: Effortless and precise human movement in real- time,” https://realmove.eu/

work page
[5]

Livehps: lidar-based scene-level human pose and shape estimation in free envi- ronment,

Y . Ren, X. Han, C. Zhao, J. Wang, L. Xu, J. Yu, and Y . Ma, “Livehps: lidar-based scene-level human pose and shape estimation in free envi- ronment,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp. 1281–1291

work page 2024
[6]

Lidarcapv2: 3d human pose estimation with human–object interaction from lidar point clouds,

J. Zhang, Q. Mao, S. Shen, C. Wen, L. Xu, and C. Wang, “Lidarcapv2: 3d human pose estimation with human–object interaction from lidar point clouds,” Pattern Recognition, vol. 156, p. 110848, 2024

work page 2024

Show all 21 references

[7]

A framework for fall detection based on openpose skeleton and lstm/gru models,

C.-B. Lin, Z. Dong, W.-K. Kuan, and Y .-F. Huang, “A framework for fall detection based on openpose skeleton and lstm/gru models,” Applied Sciences, vol. 11, no. 1, p. 329, 2020

2020
[8]

Real-time human action prediction using pose estimation with attention-based lstm network,

A. Bharathi, R. Sanku, M. Sridevi, S. Manusubramanian, and S. K. Chandar, “Real-time human action prediction using pose estimation with attention-based lstm network,” Signal, Image and Video Processing, vol. 18, no. 4, pp. 3255–3264, 2024

2024
[9]

Fall prediction based on key points of human bones,

Q. Xu, G. Huang, M. Yu, and Y . Guo, “Fall prediction based on key points of human bones,” Physica A: Statistical Mechanics and its Applications, vol. 540, p. 123205, 2020

2020
[10]

Human activity recognition: A comprehensive review,

H. Kaur, V . Rani, and M. Kumar, “Human activity recognition: A comprehensive review,”Expert Systems, vol. 41, no. 11, p. e13680, 2024

2024
[11]

Csi2image: Image reconstruction from channel state information using generative adversarial networks,

S. Kato, T. Fukushima, T. Murakami, H. Abeysekera, Y . Iwasaki, T. Fuji- hashi, T. Watanabe, and S. Saruwatari, “Csi2image: Image reconstruction from channel state information using generative adversarial networks,” IEEE Access , vol. 9, pp. 47 154–47 168, 2021

2021
[12]

Human silhouette and skeleton video synthesis through wi-fi signals,

D. Avola, M. Cascio, L. Cinque, A. Fagioli, and G. L. Foresti, “Human silhouette and skeleton video synthesis through wi-fi signals,” Interna- tional Journal of Neural Systems , vol. 32, no. 05, p. 2250015, 2022

2022
[14]

Mm-fi: Multi-modal non-intrusive 4d human dataset for versatile wireless sensing,

J. Yang, H. Huang, Y . Zhou, X. Chen, Y . Xu, S. Yuan, H. Zou, C. X. Lu, and L. Xie, “Mm-fi: Multi-modal non-intrusive 4d human dataset for versatile wireless sensing,” in Thirty-seventh Conference on Neural Information Processing Systems Datasets and Benchmarks Track , 2023. ...

2023
[15]

Skeleton-based action recog- nition with directed graph neural networks,

L. Shi, Y . Zhang, J. Cheng, and H. Lu, “Skeleton-based action recog- nition with directed graph neural networks,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2019, pp. 7912–7921

2019
[17]

From signal to image: Capturing fine-grained human poses with commodity wi-fi,

L. Guo, Z. Lu, X. Wen, S. Zhou, and Z. Han, “From signal to image: Capturing fine-grained human poses with commodity wi-fi,” IEEE Communications Letters , vol. 24, no. 4, pp. 802–806, 2019

2019
[18]

Wifall: Device-free fall detection by wireless networks,

Y . Wang, K. Wu, and L. M. Ni, “Wifall: Device-free fall detection by wireless networks,” IEEE Transactions on Mobile Computing , vol. 16, no. 2, pp. 581–594, 2016

2016
[19]

Differential channel-state-information-based human activity recognition in iot networks,

P. Khan, B. S. K. Reddy, A. Pandey, S. Kumar, and M. Youssef, “Differential channel-state-information-based human activity recognition in iot networks,” IEEE Internet of Things Journal , vol. 7, no. 11, pp. 11 290–11 302, 2020

2020
[20]

Wi-tar: Object detection system based on csi ratio,

M. Peng, B. Ge, X. Fu, and C. Kai, “Wi-tar: Object detection system based on csi ratio,” IEEE Sensors Journal , 2024

2024
[21]

Non-invasive detection of moving and stationary human with wifi,

C. Wu, Z. Yang, Z. Zhou, X. Liu, Y . Liu, and J. Cao, “Non-invasive detection of moving and stationary human with wifi,” IEEE Journal on Selected Areas in Communications , vol. 33, no. 11, pp. 2329–2342, 2015

2015

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.

[1] [13]

Metafi++: Wifi-enabled transformer-based human pose estimation for metaverse avatar simulation,

Y . Zhou, H. Huang, S. Yuan, H. Zou, L. Xie, and J. Yang, “Metafi++: Wifi-enabled transformer-based human pose estimation for metaverse avatar simulation,” IEEE Internet of Things Journal , vol. 10, no. 16, pp. 14 128–14 136, 2023

work page 2023

[2] [16]

Metafi: Device-free pose estimation via commodity wifi for metaverse avatar simulation,

J. Yang, Y . Zhou, H. Huang, H. Zou, and L. Xie, “Metafi: Device-free pose estimation via commodity wifi for metaverse avatar simulation,” in 2022 IEEE 8th World F orum on Internet of Things (WF-IoT) . IEEE, 2022, pp. 1–6

work page 2022

[3] [1]

Ultralytics yolo11,

G. Jocher and J. Qiu, “Ultralytics yolo11,” 2024. [Online]. Available: https://github.com/ultralytics/ultralytics 8

work page 2024

[4] [2]

Openpose: Realtime multi-person 2d pose estimation using part affinity fields,

Z. Cao, G. Hidalgo, T. Simon, S.-E. Wei, and Y . Sheikh, “Openpose: Realtime multi-person 2d pose estimation using part affinity fields,” IEEE transactions on pattern analysis and machine intelligence , vol. 43, no. 1, pp. 172–186, 2019

work page 2019

[5] [3]

Blazepose: On-device real-time body pose tracking,

V . Bazarevsky, I. Grishchenko, K. Raveendran, T. Zhu, F. Zhang, and M. Grundmann, “Blazepose: On-device real-time body pose tracking,” arXiv preprint arXiv:2006.10204 , 2020

arXiv 2006

[6] [4]

Real move: Effortless and precise human movement in real- time,

R. Move, “Real move: Effortless and precise human movement in real- time,” https://realmove.eu/

work page

[7] [5]

Livehps: lidar-based scene-level human pose and shape estimation in free envi- ronment,

Y . Ren, X. Han, C. Zhao, J. Wang, L. Xu, J. Yu, and Y . Ma, “Livehps: lidar-based scene-level human pose and shape estimation in free envi- ronment,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp. 1281–1291

work page 2024

[8] [6]

Lidarcapv2: 3d human pose estimation with human–object interaction from lidar point clouds,

J. Zhang, Q. Mao, S. Shen, C. Wen, L. Xu, and C. Wang, “Lidarcapv2: 3d human pose estimation with human–object interaction from lidar point clouds,” Pattern Recognition, vol. 156, p. 110848, 2024

work page 2024