pith. machine review for the scientific record. sign in

arxiv: 2605.01369 · v1 · submitted 2026-05-02 · 📡 eess.SP · cs.AI· cs.LG

Recognition: unknown

MU-SHOT-Fi: Self-Supervised Multi-User Wi-Fi Sensing with Source-free Unsupervised Domain Adaptation

Ahmed Y. Radwan , Hina Tabassum
Authors on Pith no claims yet

Pith reviewed 2026-05-09 18:30 UTC · model grok-4.3

classification 📡 eess.SP cs.AIcs.LG
keywords Wi-Fi sensingchannel state informationhuman activity recognitiondomain adaptationself-supervised learningmulti-user sensingsource-free adaptationoccupancy estimation
0
0 comments X

The pith

MU-SHOT-Fi adapts a pre-trained Wi-Fi model to new rooms and frequencies using only unlabeled target CSI and self-supervision.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces MU-SHOT-Fi as a source-free unsupervised domain adaptation method for Wi-Fi channel state information based human activity recognition in both single-user and multi-user settings. It trains a source model with permutation-invariant set prediction via Hungarian matching, then in the target domain freezes the classifier and adapts the backbone through occupancy-weighted information maximization plus binary rotation prediction. These components maintain performance on exact activity labels and occupancy counts across cross-environment, cross-frequency, and cross-orientation shifts without access to source data or any target labels. A sympathetic reader would care because privacy rules often block sharing of labeled source recordings, yet real deployments must handle changing rooms, devices, and user counts. If the method holds, it removes the need for fresh labeled collections in each new physical space.

Core claim

MU-SHOT-Fi performs source training with permutation-invariant set prediction and Hungarian matching to handle variable user counts, then applies frozen-classifier adaptation in the target domain. It stabilizes adaptation via occupancy-weighted information maximization that applies diversity pressure only to likely-occupied time slots while excluding the dominant class from marginal entropy, together with binary rotation prediction that exploits CSI frequency-time structure for domain-invariant features. The single-user variant replaces occupancy weighting with standard information maximization and adds contrastive predictive coding for temporal consistency. Experiments across the WiMANS and

What carries the argument

MU-SHOT-Fi's frozen-classifier backbone adaptation driven by occupancy-weighted information maximization and binary rotation prediction self-supervision.

If this is right

  • Multi-user exact-activity classification accuracy is restored under large domain shifts.
  • Occupancy estimation remains accurate without collapse to the dominant class.
  • The same pipeline works for single-user cases after swapping in standard information maximization and contrastive predictive coding.
  • Performance holds across cross-environment, cross-frequency, cross-orientation, and combined shifts on standard CSI datasets.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The approach could extend to other RF sensing tasks such as gesture or gait recognition where labeled data collection is costly.
  • If the initial source model is trained on highly diverse environments, the adaptation step may become unnecessary in many practical cases.
  • Deployment on edge hardware would allow on-device self-supervision loops that continuously track slow environmental drift.

Load-bearing premise

A pre-trained source model already encodes sufficiently general features so that self-supervision alone can realign them to a new domain without any target labels.

What would settle it

Train the source model on a restricted activity set in one room, then measure whether MU-SHOT-Fi recovers exact multi-user activity labels when the target domain contains previously unseen activity combinations and strong interference.

Figures

Figures reproduced from arXiv: 2605.01369 by Ahmed Y. Radwan, Hina Tabassum.

Figure 1
Figure 1. Figure 1: Architecture of the proposed MU-SHOT-Fi source-free unsupervised domain adaptation framework for multi-user Wi-Fi view at source ↗
Figure 2
Figure 2. Figure 2: Architecture of the proposed SU-SHOT-Fi source-free unsupervised domain adaptation framework for single-user Wi-Fi view at source ↗
Figure 3
Figure 3. Figure 3: Per-class F1-scores under Cross-Frequency shift view at source ↗
read the original abstract

Deep learning has been widely adopted for WiFi CSI-based human activity recognition (HAR) due to its ability to learn spatio-temporal features in a privacy-preserving and cost-effective manner. However, DL-based models generalize poorly across environments, a challenge amplified in multi-user settings where overlapping activities cause CSI entanglement and domain shifts. Practical deployments often limit access to labeled source data due to privacy constraints, motivating source-free adaptation using only unlabeled target-domain CSI and a pre-trained source model. In this paper, we propose MU-SHOT-Fi, a source-free unsupervised domain adaptation framework for single- and multi-user Wi-Fi sensing. MU-SHOT-Fi employs permutation-invariant set prediction with Hungarian matching during source training, followed by frozen-classifier backbone adaptation in the target domain. To enable stable adaptation without labels, we introduce occupancy-weighted information maximization that prevents model collapse by focusing diversity regularization on likely-occupied slots while excluding the dominant class from marginal entropy. Additionally, we employ binary rotation prediction as spatial self-supervision that exploits CSI frequency-time structure to learn domain-invariant features. For single-user scenarios, we introduce SU-SHOT-Fi by replacing occupancy weighting with standard information maximization and incorporating contrastive predictive coding to exploit temporal consistency. Extensive experiments on the WiMANS and Widar 3.0 datasets across cross-environment, cross-frequency, cross-orientation, and combined domain shifts demonstrate that MU-SHOT-Fi effectively recovers multi-user exact-activity classification performance under large domain shifts while maintaining accurate occupancy estimation and preventing collapse toward dominant classes.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The manuscript proposes MU-SHOT-Fi, a source-free unsupervised domain adaptation framework for single- and multi-user Wi-Fi CSI-based human activity recognition. It pre-trains a permutation-invariant set prediction model using Hungarian matching on the source domain, then performs target-only backbone adaptation with a frozen source classifier. The target adaptation uses occupancy-weighted information maximization (with dominant-class exclusion from marginal entropy) plus binary rotation prediction as self-supervision to prevent collapse and learn invariant features; a single-user variant (SU-SHOT-Fi) replaces these with standard information maximization and contrastive predictive coding. Experiments on WiMANS and Widar 3.0 claim effective recovery of exact-activity classification under cross-environment, cross-frequency, cross-orientation, and combined shifts while preserving occupancy estimation.

Significance. If the empirical claims hold with rigorous validation, the work would be significant for practical Wi-Fi sensing deployments constrained by privacy (no source data access) and multi-user entanglement. The combination of set prediction, occupancy-aware regularization, and rotation-based self-supervision targets specific failure modes in UDA for entangled signals, extending single-user methods to more realistic scenarios.

major comments (3)
  1. [§3.2] §3.2 (adaptation framework): The central claim that frozen-classifier adaptation recovers exact multi-user activity labels rests on the assumption that occupancy-weighted information maximization and binary rotation prediction can realign source features under CSI superposition and large shifts (e.g., cross-frequency). However, no feature alignment metrics, t-SNE visualizations, or analysis of how rotation prediction disentangles user-specific paths are provided, leaving the weakest assumption untested.
  2. [§4] §4 (experiments): The abstract and results claim performance recovery and prevention of dominant-class collapse, yet no quantitative tables, error bars, ablation studies on the occupancy weighting coefficient or rotation loss weight, or statistical tests are referenced. This undermines evaluation of whether the gains are reliable or merely post-hoc tuning artifacts on the same datasets used for design choices.
  3. [§3.2] §3.2 (occupancy weighting): The exclusion of the dominant class from marginal entropy and the use of occupancy estimates from the unadapted model are load-bearing for stable adaptation, but no justification, sensitivity analysis, or ablation demonstrates these choices are necessary versus standard information maximization.
minor comments (2)
  1. [Abstract] The abstract states 'extensive experiments demonstrate...' but contains no numerical results, which is atypical and reduces immediate impact.
  2. [§4] Ensure all free parameters (occupancy weighting coefficient, rotation prediction loss weight) and their selection procedure are explicitly reported with values used for each dataset/shift.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive and detailed feedback, as well as the positive assessment of the potential significance of MU-SHOT-Fi for practical, privacy-constrained Wi-Fi sensing deployments. We address each major comment point by point below and will make the corresponding revisions to strengthen the manuscript.

read point-by-point responses

  1. Referee: [§3.2] §3.2 (adaptation framework): The central claim that frozen-classifier adaptation recovers exact multi-user activity labels rests on the assumption that occupancy-weighted information maximization and binary rotation prediction can realign source features under CSI superposition and large shifts (e.g., cross-frequency). However, no feature alignment metrics, t-SNE visualizations, or analysis of how rotation prediction disentangles user-specific paths are provided, leaving the weakest assumption untested.

    Authors: We agree that direct evidence of feature realignment would strengthen support for the adaptation claims. In the revised manuscript, we will add t-SNE visualizations of source and target features before and after adaptation for representative cross-environment and cross-frequency shifts. We will also report quantitative alignment metrics such as Maximum Mean Discrepancy (MMD) between source and adapted target embeddings. Additionally, we will include a discussion analyzing how the binary rotation prediction self-supervision exploits CSI frequency-time structure to encourage domain-invariant representations, thereby aiding disentanglement of superimposed user signals; this will be tied to observed improvements in multi-user exact-activity accuracy. revision: yes

  2. Referee: [§4] §4 (experiments): The abstract and results claim performance recovery and prevention of dominant-class collapse, yet no quantitative tables, error bars, ablation studies on the occupancy weighting coefficient or rotation loss weight, or statistical tests are referenced. This undermines evaluation of whether the gains are reliable or merely post-hoc tuning artifacts on the same datasets used for design choices.

    Authors: Section 4 already presents quantitative tables reporting exact-activity classification accuracies and occupancy estimation errors for MU-SHOT-Fi, SU-SHOT-Fi, and baselines across all evaluated domain shifts on WiMANS and Widar 3.0. To further address reliability concerns, the revision will add error bars (standard deviations over multiple random seeds), ablation studies systematically varying the occupancy weighting coefficient and rotation loss weight, and statistical significance tests (e.g., paired t-tests) comparing our method against baselines. These additions will confirm that performance gains are consistent and not artifacts of post-hoc tuning. revision: yes

  3. Referee: [§3.2] §3.2 (occupancy weighting): The exclusion of the dominant class from marginal entropy and the use of occupancy estimates from the unadapted model are load-bearing for stable adaptation, but no justification, sensitivity analysis, or ablation demonstrates these choices are necessary versus standard information maximization.

    Authors: We will expand §3.2 with a detailed justification explaining how occupancy weighting and dominant-class exclusion from marginal entropy mitigate collapse in multi-user settings with imbalanced occupancy distributions, while leveraging unadapted occupancy estimates for stable target-only adaptation. The revision will include a sensitivity analysis (performance curves over a range of weighting coefficients) and an ablation study comparing the proposed occupancy-weighted information maximization against standard information maximization. These will empirically demonstrate the necessity of the design choices for preventing dominant-class collapse. revision: yes

Circularity Check

0 steps flagged

Novel self-supervised losses and adaptation logic are empirically validated without reducing to definitional equivalence or self-citation tautology

full rationale

The paper's chain begins with source pre-training via permutation-invariant set prediction and Hungarian matching, followed by target-domain frozen-classifier adaptation using occupancy-weighted information maximization (to avoid collapse) and binary rotation prediction for domain-invariant features. These components are introduced as original proposals and evaluated empirically on WiMANS and Widar 3.0 under multiple domain shifts. No equation or claim reduces by construction to its own inputs (e.g., no fitted parameter renamed as prediction, no self-definitional loop, and no load-bearing uniqueness theorem imported solely via author self-citation). The central performance-recovery claim remains an empirical assertion supported by experimental results rather than a tautology, though effectiveness is demonstrated on the same benchmark datasets used for tuning.

Axiom & Free-Parameter Ledger

2 free parameters · 2 axioms · 0 invented entities

Only the abstract is available; the ledger therefore reflects components explicitly named in the abstract rather than full derivation details.

free parameters (2)
  • occupancy weighting coefficient
    Weight applied to information-maximization loss on likely-occupied slots; value chosen to balance diversity and prevent collapse.
  • rotation prediction loss weight
    Hyperparameter balancing spatial self-supervision against the main adaptation objective.
axioms (2)
  • domain assumption Pre-trained source model features remain useful after freezing the classifier
    Invoked when the method performs frozen-classifier backbone adaptation in the target domain.
  • domain assumption Unlabeled target CSI contains sufficient structure for self-supervision to learn domain-invariant features
    Underlying the use of binary rotation prediction and contrastive predictive coding.

pith-pipeline@v0.9.0 · 5582 in / 1563 out tokens · 23307 ms · 2026-05-09T18:30:53.813995+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

44 extracted references · 3 canonical work pages · 1 internal anchor

  1. [1]

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

    A. Y . Radwan, M. Yildirim, N. Hasanzadeh, H. Tabassum, and S. Valaee, “A tutorial-cum-survey on self-supervised learning for wi-fi sensing: Trends, challenges, and outlook,”IEEE Commun. Surveys & Tut., 2025

    2025

  2. [2]

    Rscnet: Dynamic csi compression for cloud-based wifi sensing,

    B. Barahimi, H. Singh, H. Tabassum, O. Waqar, and M. Omer, “Rscnet: Dynamic csi compression for cloud-based wifi sensing,” inICC 2024- IEEE International Conference on Communications. IEEE, 2024, pp. 4179–4184

    2024

  3. [3]

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

    J. Ding, Y . Wang, Q. Xie, and J. Niu, “A multiple wifi sensors assisted human activity recognition scheme for smart home,”IEEE Sensors Journal, 2024

    2024

  4. [4]

    Environment-Robust WiFi-Based Human Activity Recognition Using Enhanced CSI and Deep Learning,

    Z. Shi, Q. Cheng, J. A. Zhang, and R. Y . Da Xu, “Environment-Robust WiFi-Based Human Activity Recognition Using Enhanced CSI and Deep Learning,”IEEE Internet of Things Jrnl., vol. 9, no. 24, pp. 24 643– 24 654, 2022

    2022

  5. [5]

    Deep transfer learning for WiFi localization,

    P. Li, H. Cui, A. Khan, U. Raza, R. Piechocki, A. Doufexi, and T. Farnham, “Deep transfer learning for WiFi localization,” in2021 IEEE Radar Conf. (RadarConf21). IEEE, 2021, pp. 1–5

    2021

  6. [6]

    Wibot! in-vehicle behaviour and gesture recognition using wireless network edge,

    M. Raja, V . Ghaderi, and S. Sigg, “Wibot! in-vehicle behaviour and gesture recognition using wireless network edge,” in2018 IEEE 38th International Conf. on Distributed Computing Sys. (ICDCS). IEEE, 2018, pp. 376–387

    2018

  7. [7]

    Heart rate variability extraction using com- modity Wi-Fi devices via time domain signal processing,

    I. Shirakami and T. Sato, “Heart rate variability extraction using com- modity Wi-Fi devices via time domain signal processing,” in2021 IEEE EMBS Intl. Conf. on Biomedical and Health Informatics (BHI). IEEE, 2021, pp. 1–4

    2021

  8. [8]

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

    D. Wu, Y . Zeng, F. Zhang, and D. Zhang, “Wifi csi-based device-free sensing: from fresnel zone model to csi-ratio model,”CCF Trans. on Pervasive Computing and Interaction, vol. 4, no. 1, pp. 88–102, 2022

    2022

  9. [9]

    Estimating angle-of-arrival and time-of-flight for multi- path components using wifi channel state information,

    A. U. Ahmed, R. Arablouei, F. De Hoog, B. Kusy, R. Jurdak, and N. Bergmann, “Estimating angle-of-arrival and time-of-flight for multi- path components using wifi channel state information,”Sensors, vol. 18, no. 6, p. 1753, 2018

    2018

  10. [10]

    A survey on behavior recognition using wifi channel state information,

    S. Yousefi, H. Narui, S. Dayal, S. Ermon, and S. Valaee, “A survey on behavior recognition using wifi channel state information,”IEEE Commun. Magazine, vol. 55, no. 10, pp. 98–104, 2017

    2017

  11. [11]

    Time matters: Empirical insights into the limits and challenges of temporal generalization in csi- based wi-fi sensing,

    A. Brunello, A. Montanari, R. Montoliu, A. Moreira, N. Saccomanno, E. Sansano-Sansano, and J. Torres-Sospedra, “Time matters: Empirical insights into the limits and challenges of temporal generalization in csi- based wi-fi sensing,”Internet of Things, p. 101634, 2025

    2025

  12. [12]

    Exposing the csi: A systematic investigation of csi-based wi-fi sensing capabilities and limitations,

    M. Cominelli, F. Gringoli, and F. Restuccia, “Exposing the csi: A systematic investigation of csi-based wi-fi sensing capabilities and limitations,” in2023 IEEE International Conf. on Pervasive Computing and Commun. (PerCom). IEEE, 2023, pp. 81–90

    2023

  13. [13]

    Wifi sensing with channel state information: A survey,

    Y . Ma, G. Zhou, S. Wang, H. Zhao, and W. Jung, “Wifi sensing with channel state information: A survey,”ACM Computing Surveys, vol. 52, no. 3, pp. 46:1–46:36, 2019

    2019

  14. [14]

    Wifi csi based passive human activity recognition using attention based blstm,

    Z. Chen, L. Zhang, C. Jiang, Z. Cao, and W. Cui, “Wifi csi based passive human activity recognition using attention based blstm,”IEEE Trans. on Mobile Computing, vol. 18, no. 11, pp. 2714–2724, 2018

    2018

  15. [15]

    Context-aware predictive coding: A representation learning framework for wifi sensing,

    B. Barahimi, H. Tabassum, M. Omer, and O. Waqar, “Context-aware predictive coding: A representation learning framework for wifi sensing,” IEEE Open Journal of the Communications Society, vol. 5, pp. 6119– 6134, 2024

    2024

  16. [16]

    Commodity WiFi sensing in ten years: Status, challenges, and opportunities,

    S. Tan, Y . Ren, J. Yang, and Y . Chen, “Commodity WiFi sensing in ten years: Status, challenges, and opportunities,”IEEE Internet of Things Journal, vol. 9, no. 18, pp. 17 832–17 843, 2022. 16

    2022

  17. [17]

    Wimans: A benchmark dataset for wifi-based multi-user activity sensing,

    S. Huang, K. Li, D. You, Y . Chen, A. Lin, S. Liu, X. Li, and J. A. McCann, “Wimans: A benchmark dataset for wifi-based multi-user activity sensing,” inEuropean Conf. on Computer Vision. Springer, 2024, pp. 72–91

    2024

  18. [18]

    Multitrack: Multi-user tracking and activity recognition using commodity wifi,

    S. Tan, L. Zhang, Z. Wang, and J. Yang, “Multitrack: Multi-user tracking and activity recognition using commodity wifi,” inProceedings of the 2019 CHI Conf. on Human Factors in Computing Sys., 2019, pp. 1–12

    2019

  19. [19]

    Tensorbeat: Tensor decomposition for monitoring multiperson breathing beats with commodity wifi,

    X. Wang, C. Yang, and S. Mao, “Tensorbeat: Tensor decomposition for monitoring multiperson breathing beats with commodity wifi,”ACM Trans. on Intelligent Sys. and Technology (TIST), vol. 9, no. 1, pp. 1–27, 2017

    2017

  20. [20]

    Multisensex: A sus- tainable solution for multi-human activity recognition and localization in smart environments,

    H. Rizk, A. Elmogy, M. Rihan, and H. Yamaguchi, “Multisensex: A sus- tainable solution for multi-human activity recognition and localization in smart environments,”AI, vol. 6, no. 1, p. 6, 2025

    2025

  21. [21]

    Data augmentation techniques for cross- domain wifi csi-based human activity recognition,

    J. Strohmayer and M. Kampel, “Data augmentation techniques for cross- domain wifi csi-based human activity recognition,” inIFIP International Conf. on Artificial Intelligence Applications and Innovations. Springer, 2024, pp. 42–56

    2024

  22. [22]

    Unsupervised domain adaptation for rf-based gesture recognition,

    B.-B. Zhang, D. Zhang, Y . Li, Y . Hu, and Y . Chen, “Unsupervised domain adaptation for rf-based gesture recognition,”IEEE Internet of Things Journal, vol. 10, no. 23, pp. 21 026–21 038, 2023

    2023

  23. [23]

    Fidora: Robust wifi-based indoor localization via unsupervised domain adaptation,

    X. Chen, H. Li, C. Zhou, X. Liu, D. Wu, and G. Dudek, “Fidora: Robust wifi-based indoor localization via unsupervised domain adaptation,” IEEE Internet of Things Journal, vol. 9, no. 12, pp. 9872–9888, 2022

    2022

  24. [24]

    Robust indoor localization in dynamic environments: A multi-source unsupervised domain adaptation frame- work,

    J. Jiao, X. Wang, and C. Han, “Robust indoor localization in dynamic environments: A multi-source unsupervised domain adaptation frame- work,”arXiv preprint arXiv:2502.07246, 2025

    work page arXiv 2025

  25. [25]

    Do we really need to access the source data? source hypothesis transfer for unsupervised domain adaptation,

    J. Liang, D. Hu, and J. Feng, “Do we really need to access the source data? source hypothesis transfer for unsupervised domain adaptation,” in International Conf. on machine learning. PMLR, 2020, pp. 6028–6039

    2020

  26. [26]

    A comprehensive survey on source-free domain adaptation,

    J. Li, Z. Yu, Z. Du, L. Zhu, and H. T. Shen, “A comprehensive survey on source-free domain adaptation,”IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 46, no. 8, pp. 5743–5762, 2024

    2024

  27. [27]

    Wi-sfdagr: Wifi-based cross-domain gesture recognition via source-free domain adaptation,

    H. Yan, X. Zhang, J. Huang, Y . Feng, M. Li, A. Wang, W. Ou, H. Wang, and Z. Liu, “Wi-sfdagr: Wifi-based cross-domain gesture recognition via source-free domain adaptation,”IEEE Internet of Things Journal, 2025

    2025

  28. [28]

    Towards environment-independent behavior-based user authentication using wifi,

    C. Shi, J. Liu, N. Borodinov, B. Leao, and Y . Chen, “Towards environment-independent behavior-based user authentication using wifi,” in2020 IEEE 17th International Conf. on Mobile Ad Hoc and Sensor Sys. (MASS). IEEE, 2020, pp. 666–674

    2020

  29. [29]

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

    Y . Zhang, Y . Zheng, K. Qian, G. Zhang, Y . Liu, C. Wu, and Z. Yang, “Widar3.0: Zero-effort cross-domain gesture recognition with wi-fi,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 44, no. 11, pp. 8671–8688, 2022

    2022

  30. [30]

    Channel state information prediction for 5g wireless communications: A deep learning approach,

    C. Luo, J. Ji, Q. Wang, X. Chen, and P. Li, “Channel state information prediction for 5g wireless communications: A deep learning approach,” IEEE Trans. on network science and engineering, vol. 7, no. 1, pp. 227–236, 2018

    2018

  31. [31]

    A survey on wifi-based human identification: Scenarios, challenges, and current solutions,

    Z. Wei, W. Chen, S. Ning, W. Lin, N. Li, B. Lian, X. Sun, and J. Zhao, “A survey on wifi-based human identification: Scenarios, challenges, and current solutions,”ACM Trans. on Sensor Networks, vol. 21, no. 1, pp. 1–32, 2025

    2025

  32. [32]

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

    N. Hasanzadeh and S. Valaee, “Enhancing generalization in human activity recognition through improved wi-fi channel state information phase processing and antenna pair selection,” in2024 IEEE 34th International Workshop on Machine Learning for Signal Processing (MLSP). IEEE, 2024, pp. 1–6

    2024

  33. [33]

    Wiai-id: Wi-fi-based domain adaptation for appearance-independent passive person identification,

    Y . Liang, W. Wu, H. Li, F. Han, Z. Liu, P. Xu, X. Lian, and X. Chen, “Wiai-id: Wi-fi-based domain adaptation for appearance-independent passive person identification,”IEEE Internet of Things Journal, vol. 11, no. 1, pp. 1012–1027, 2023

    2023

  34. [34]

    Source data-absent unsupervised domain adaptation through hypothesis transfer and labeling transfer,

    J. Liang, D. Hu, Y . Wang, R. He, and J. Feng, “Source data-absent unsupervised domain adaptation through hypothesis transfer and labeling transfer,”IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 44, no. 11, pp. 8602–8617, 2021

    2021

  35. [35]

    Source-free unsu- pervised domain adaptation: A survey,

    Y . Fang, P.-T. Yap, W. Lin, H. Zhu, and M. Liu, “Source-free unsu- pervised domain adaptation: A survey,”Neural Networks, vol. 174, p. 106230, 2024

    2024

  36. [36]

    End-to-end object detection with transformers,

    N. Carion, F. Massa, G. Synnaeve, N. Usunier, A. Kirillov, and S. Zagoruyko, “End-to-end object detection with transformers,” in European Conf. on computer vision. Springer, 2020, pp. 213–229

    2020

  37. [37]

    The hungarian method for the assignment problem,

    H. W. Kuhn, “The hungarian method for the assignment problem,”Naval research logistics quarterly, vol. 2, no. 1-2, pp. 83–97, 1955

    1955

  38. [38]

    Learning imbal- anced datasets with label-distribution-aware margin loss,

    K. Cao, C. Wei, A. Gaidon, N. Arechiga, and T. Ma, “Learning imbal- anced datasets with label-distribution-aware margin loss,”Advances in neural information processing systems, vol. 32, 2019

    2019

  39. [39]

    Unsupervised representation learning by predicting image rotations

    S. Gidaris, P. Singh, and N. Komodakis, “Unsupervised repre- sentation learning by predicting image rotations,”arXiv preprint arXiv:1803.07728, 2018

    work page arXiv 2018

  40. [40]

    Bootstrap your own latent-a new approach to self-supervised learning,

    J.-B. Grill, F. Strub, F. Altch ´e, C. Tallec, P. Richemond, E. Buchatskaya, C. Doersch, B. Avila Pires, Z. Guo, M. Gheshlaghi Azaret al., “Bootstrap your own latent-a new approach to self-supervised learning,” Advances in neural information processing Sys., vol. 33, pp. 21 271– 21 284, 2020

    2020

  41. [41]

    Representation Learning with Contrastive Predictive Coding

    A. v. d. Oord, Y . Li, and O. Vinyals, “Representation learning with contrastive predictive coding,”arXiv preprint arXiv:1807.03748, 2018

    work page internal anchor Pith review Pith/arXiv arXiv 2018

  42. [42]

    A survey on wi-fi sensing generalizability: Taxonomy, techniques, datasets, and future research prospects,

    F. Wang, T. Zhang, W. Xi, H. Ding, G. Wang, D. Zhang, Y . Cui, F. Liu, J. Han, J. Xuet al., “A survey on wi-fi sensing generalizability: Taxonomy, techniques, datasets, and future research prospects,”IEEE Communications Surveys & Tutorials, 2026

    2026

  43. [43]

    Mixmatch: A holistic approach to semi-supervised learning,

    D. Berthelot, N. Carlini, I. Goodfellow, N. Papernot, A. Oliver, and C. A. Raffel, “Mixmatch: A holistic approach to semi-supervised learning,” Advances in neural information processing Sys., vol. 32, 2019

    2019

  44. [44]

    A theory of learning from different domains,

    S. Ben-David, J. Blitzer, K. Crammer, A. Kulesza, F. Pereira, and J. W. Vaughan, “A theory of learning from different domains,”Machine learning, vol. 79, no. 1-2, pp. 151–175, 2010. Ahmed Radwanreceived the M.Sc. degree in com- puter science from York University, Toronto, ON, Canada, in 2026, under the supervision of Dr. H. Tabassum at the Next Generation...

    2010