Pith. sign in

REVIEW 4 major objections 5 minor 32 references

Tracking UWB Devices Through Radio Frequency Fingerprinting Is Possible

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read Ultra-wideband radio signals carry device-specific hardware fingerprints that deep learning can extract with over 99% accuracy in stable conditions and with useful accuracy at unseen locations, making both physical-layer authentication…

desk verdict First systematic UWB RFF study with open data, but the headline generalization figures rest on only three held-out locations. read the letter →

arxiv 2501.04401 v1 pith:327MVBGI submitted 2025-01-08 cs.LG cs.ITcs.NImath.IT

Thibaud Ardoin , Niklas Pauli , Benedikt Gro{ss} , Mahsa Kholghi , Khan Reaz , Gerhard Wunder This is my paper · ORCID
classification cs.LGcs.ITcs.NImath.IT
keywords ultra-widebandradiofrequencyfingerprintingdeeplearningdevicere-identificationopen-setclassificationphysicallayersecuritychannelimpulseresponseArcFaceloss
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 asks whether an ultra-wideband (UWB) radio signal, the technology behind centimeter-level indoor localization in recent phones, carries a hardware fingerprint unique enough to identify the transmitter. It answers yes on a dataset of 1.5 million measurements from 13 off-the-shelf transmitter boards under controlled position changes: a deep learning pipeline tells the transmitters apart with over 99% accuracy when training and test locations coincide. The point is not only that UWB fingerprints exist, but that they survive environmental change well enough to matter: an F1 score near 65% and an AUROC of 0.92 at locations never seen in training, rising to 76.9% when several signal samples are combined. If the result holds, physical-layer authentication of UWB devices becomes practical, and so does covert tracking of individual devices.

What carries the argument

The load-bearing machinery is a representation-learning pipeline that turns each 250-sample complex channel-impulse-response measurement into a spectrogram via short-time Fourier transform, normalizes amplitude, time-aligns the first pulse, and trains a Vision Transformer to embed each device in a feature space where measurements of the same device cluster. An additive angular margin loss (ArcFace) enforces intra-class compactness and inter-class separation, which lets the system operate open-set: after training, stored reference embeddings stand for known identities, and new measurements are matched against them by distance. A CNN baseline provides the comparison that isolates what the transformer and loss add. In the multi-sample variant, several measurements are concatenated as input, allowing the model to pool evidence across locations and raising the unseen-location F1 to 76.9%.

What would settle it

Repeat the closed-set scenario with training and test locations completely disjoint, or train a control model to predict location rather than device identity; if device accuracy collapses to chance or the location-control matches the device model, the claimed fingerprint is location leakage rather than hardware signature.

Watch

Extended reading notes

Core claim

The work claims to be the first radio-frequency fingerprinting study for UWB, and it supports the claim with a controlled, publicly released dataset. Thirteen identical transmitter boards and one receiver were mounted on a rotating rail, producing complex channel-impulse-response measurements at fifty precise locations on two days and at two distances. A convolutional baseline and a Vision Transformer with ArcFace loss are trained on short-time Fourier transform representations of the signals, then evaluated in four scenarios of increasing difficulty. In the closed-set scenario the transformer reaches 99.9% CF1 and 0.99 AUROC; moving to unseen locations drops CF1 to 64.6% but keeps AUROC at 0.92; open-set re-identification of withheld devices reaches 34.9% CF1 and 0.76 AUROC; and a completely different day, distance, and room still yields accuracy about twice as high as random. The conclusion is that device-specific hardware signatures are present and recoverable in raw UWB signals, and that limited generalization is primarily a data-diversity problem rather than the absence of a fingerprint.

Load-bearing premise

The load-bearing premise is that the controlled setup isolates hardware fingerprints from location effects, since in the closed-set scenario the training and test measurements come from the same fifty positions; if the model is instead exploiting location-specific channel signatures, the high accuracy would not mean what the paper claims.

Editorial extensions

If this is right

  • UWB physical-layer authentication is feasible: a receiver can check a tag's identity from the shape of its pulses without decrypting or modifying the data.
  • UWB tracking is a realistic privacy risk: any sniffer that can record channel impulse responses can re-identify a tag later, even in a different room, without user consent.
  • Collecting several sub-nanosecond samples per identification event is a practical way to raise reliability, since multi-sample input lifts unseen-location F1 from about 65% to 77%.
  • Deployment-grade reliability requires much larger, more diverse public datasets: accuracy drops sharply when day, distance, or room changes.
  • Physical tampering with the antenna degrades the fingerprint but does not erase it: with glue on the antenna, a linear classifier in the trained feature space drops to 44% CF1, then returns to 99% after the glue is removed.

Reading between the lines

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

  • Inference: The near-perfect closed-set score is likely inflated because training and test share the same fifty locations, so the model could be reading position-specific channel signatures rather than hardware identity alone; a strictly disjoint-location test would settle how much hardware contributes.
  • Inference: The same learned embedding could be repurposed to infer location from a single device, since it demonstrably carries enough environmental structure to distinguish positions; the paper does not explore this privacy direction.
  • Inference: Swapping antennas or heating the boards would test whether the fingerprint truly originates in the chip or partly in the antenna and cabling; if the fingerprint tracks the antenna, the hardware-stability boundary is weaker than claimed.
  • Inference: Following the paper's data-diversity argument, a multi-environment, multi-identity UWB corpus of face-recognition scale should push open-set re-identification well above the current 0.76 AUROC, which would strengthen the tracking risk even as it improves authentication.
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

4 major / 5 minor

Summary. The paper investigates whether radio frequency fingerprinting (RFF) can be applied to ultra-wideband (UWB) signals, motivated by both physical-layer security and privacy-tracking implications. The authors collected a controlled dataset from 13 off-the-shelf UWB transmitters plus one receiver, with 50 rotated emitter locations and 2000 measurements per location, and published it open-source. They propose a deep learning pipeline based on a Vision Transformer (ViT) with ArcFace loss, compare it to a CNN baseline, and evaluate four scenarios of increasing generalization difficulty: same-location closed-set, unseen-location closed-set, open-set re-identification, and out-of-distribution environment. They report 99.9% CF1 in Scenario 1, 64.6% CF1 in Scenario 2, 0.76 AUROC in Scenario 3, and 14.6% CF1 in Scenario 4, and further show that concatenating multiple samples can raise the Scenario 2 CF1 to 76.9%.

Significance. If the quantitative claims stand, this is a valuable first demonstration that UWB devices can be fingerprinted from raw channel impulse responses, with potential consequences for both security and privacy. The controlled dataset design and the open-set evaluation protocol are useful contributions to the RFF literature, and the introduction of transformer-based architectures plus ArcFace to UWB RFF is a reasonable novelty. The paper explicitly shares code and data, and the honesty about limitations (e.g., the conclusion's admission that proving absence of bias is challenging) is a strength. However, the central feasibility claim relies crucially on the cross-location generalization results, and those results currently rest on a very small number of independent test conditions.

major comments (4)
  1. [Section II-C and Table I, Scenario 1] The train and test sets in Scenario 1 share the same 50 locations, so the 99.9% CF1 and 0.99 AUROC could be inflated by the model learning location-specific channel signatures rather than device hardware fingerprints. The paper's statement in Section II-A that 'all other setup variables were intentionally kept constant' does not rule out location overlap as a confound. I request an explicit acknowledgment of this confound and, ideally, an analysis that quantifies how much of the classification performance survives when train and test locations are disjoint, since that is the load-bearing evidence for device-specific hardware features.
  2. [Section IV, Table I and Table II] The cross-location generalization claim in Scenario 2 rests on only three held-out locations. With 13 devices, this gives 39 independent device-location test conditions; the 2000 repeated measurements per condition are pseudoreplicates of the same geometry and do not add independent evidence. The reported 64.6% CF1 and 0.92 AUROC are point estimates with no per-location breakdown, confidence intervals, or leave-one-location-out cross-validation. A model could exceed chance by exploiting a single test location whose multipath signature correlates with device identity. Please report per-location results, bootstrap confidence intervals over locations, and a leave-one-location-out analysis to demonstrate that the result is stable across locations.
  3. [Table I] Scenarios 2–4 have no reported random baseline. The random projection baseline is given only for Scenario 1 (7.7% CF1, 0.5 AUROC), but in an open-set and out-of-distribution setting the chance-level performance can differ. In particular, Section IV-B states that Scenario 4 is 'almost twice as good as random,' but no random baseline is shown for that scenario. Please compute and report a random or majority-class baseline for each scenario over the same test distribution, so that the reader can assess the magnitude of the improvement above chance.
  4. [Section II-C and Table I, Scenario 3] The open-set evaluation in Scenario 3 holds out only three device IDs and three locations. The AUROC of 0.76 is thus based on very few enrollment/query conditions, and the conclusion that 'the limitation does not lie in the RFF extraction itself, but rather in the consistency of that extraction when compared across different locations' is too strong for this sample size. I recommend either expanding the number of held-out devices and locations, providing a per-query breakdown, or tempering the conclusion to reflect the limited statistical support.
minor comments (5)
  1. [Section IV-A] The sentence 'To summarise the ROC curve in a single interpretable value, we compute the the Area Under the ROC curve (AUROC)' contains a duplicated 'the'.
  2. [Table II caption] The caption uses the abbreviations Sl, Dl, V, and CI but does not fully expand them in the caption text; please define 'Same location' and 'Different location', and clarify that V is voting and CI is concatenated input, directly in the caption.
  3. [Equation (2)] The summation index in the denominator is written as 'j=1,j≠yi' which is nonstandard; please clarify that the sum runs over all classes except the ground-truth class, or use a cleaner notation such as 'j≠yi'.
  4. [Section IV-B] The phrase 'with a AUROC of 0.76' should be 'with an AUROC of 0.76'.
  5. [Section II-B] The STFT formula would be clearer if the window length and hop size R were explicitly defined in the text just before Eq. (1), rather than only in the architecture description.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: all key results are measured on held-out test data, and no parameter is fitted to the test labels or renamed as a prediction.

full rationale

The paper's derivation chain is empirically self-contained rather than circular. The central claims are evaluated in four scenarios, each with explicitly defined train/test splits: Scenario 1 shares locations but the test set is still a held-out set of measurements; Scenario 2 excludes three specified test locations for known device IDs; Scenario 3 excludes three device IDs and three locations for open-set evaluation; Scenario 4 uses data from a different day, distance, and room position. Hyperparameters are optimized by grid search on training/validation data, not on the test sets. The open-set reference embedding in Scenario 3 is computed from one unseen test position, which is a standard enrollment protocol and does not make the queries equivalent to the enrollment by construction. The paper does not invoke a self-citation chain as the load-bearing justification for its feasibility claim; citations to prior work are used only for architectural inspiration and baselines, and the only author-generated citation is the openly published dataset, which is not used to define the result. Statistical concerns about pseudoreplication, the small number of held-out locations, and possible location-to-device confounding are real validity limitations, but they are not circularity: they concern whether the measured accuracy reflects hardware fingerprints rather than whether the reported numbers are logically forced by the inputs. No equation is defined in terms of the target result, and no fitted parameter is renamed as a prediction. Therefore the appropriate circularity score is 0.

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

The paper introduces no new theoretical entities; its contribution is empirical. The free parameters are standard ML hyperparameters and unspecified STFT settings. The central claim rests on the domain assumption that hardware fingerprints exist and can be isolated from environmental factors.

free parameters (4)
  • ArcFace margin m = 0.1
    Chosen by grid search for intra-class compactness; not justified analytically.
  • ArcFace scale s = 64
    Chosen by grid search; adjusts softmax output.
  • ViT architecture hyperparameters (layers, heads, hidden, latent) = 1 layer, 6 heads, hidden 32, latent 192
    Determined by empirical grid search; described in Fig. 5.
  • STFT parameters (window length, hop size) = Not stated in text
    Equation (1) gives generic STFT; the actual N and R used to produce 32x32 inputs are unspecified.
assumptions (4)
  • domain assumption Hardware variations produce persistent, device-specific signal features in UWB transceivers.
    Assumed in Introduction and Sec. III-A; underpins the entire fingerprinting concept.
  • domain assumption All non-location variables (channel, orientation, receiver, room temperature within day) are constant across the dataset.
    Stated in Sec. II-A; if false, the model could latch onto environmental cues.
  • standard math The cross-entropy training objective with a linear projection yields a feature space where same-device features cluster.
    Standard deep metric learning assumption; not proven here.
  • standard math The STFT and ArcFace loss equations are correct and applicable to the discretized signals.
    Equations (1) and (2); standard signal processing and loss formulations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Tracking UWB Devices Through Radio Frequency Fingerprinting Is Possible." pith.science (2026). https://pith.science/paper/327MVBGI

@misc{pith2026250104401,
  author       = {Pith},
  title        = {Pith review of: Tracking UWB Devices Through Radio Frequency Fingerprinting Is Possible},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/327MVBGI}},
  note         = {Machine review of arXiv:2501.04401}
}
read the original abstract

Ultra-wideband (UWB) is a state-of-the-art technology designed for applications requiring centimeter-level localization. Its widespread adoption by smartphone manufacturer naturally raises security and privacy concerns. Successfully implementing Radio Frequency Fingerprinting (RFF) to UWB could enable physical layer security, but might also allow undesired tracking of the devices. The scope of this paper is to explore the feasibility of applying RFF to UWB and investigates how well this technique generalizes across different environments. We collected a realistic dataset using off-the-shelf UWB devices with controlled variation in device positioning. Moreover, we developed an improved deep learning pipeline to extract the hardware signature from the signal data. In stable conditions, the extracted RFF achieves over 99% accuracy. While the accuracy decreases in more changing environments, we still obtain up to 76% accuracy in untrained locations.

Figures

Figures reproduced from arXiv: 2501.04401 by the authors.

Figure 1
Figure 1. Experimental setup. UWB boards clipped in the 3D [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Pipeline of the RFF extraction system through representation learning [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. The evaluation scenarios with their degree of com [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: t-SNE visualisation of the data before and after the [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 5
Figure 5. Figure 5: ViT architecture with parameters [PITH_FULL_IMAGE:figures/full_fig_p004_5.png]
Figure 6
Figure 6. Figure 6: Confusion matrix on 9300 samples of scenario 2 test [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]
Figure 7
Figure 7. Figure 7: UWB boards with and without glued antenna [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

32 extracted references · 22 canonical work pages

  1. [1]

    Ultra-Wideband Technology: Characteristics, Applications and Challenges

    C. Zheng, Y . Ge, and A. Guo, “Ultra-wideband technology: Characteristics, applications and challenges,” 2023. [Online]. Available: https://arxiv.org/abs/2307.13066

    work page Pith review arXiv 2023

  2. [2]

    Radio frequency fingerprint identification for internet of things: A survey,

    L. Xie, L. Peng, J. Zhang, and A. Hu, “Radio frequency fingerprint identification for internet of things: A survey,” Security and Safety , vol. 3, p. 2023022, Sep. 2023. [Online]. Available: http://dx.doi.org/10.1051/sands/2023022

    work page arXiv 2023

  3. [3]

    A comprehensive survey on radio frequency (rf) fingerprinting: Traditional approaches, deep learning, and open challenges,

    A. Jagannath, J. Jagannath, and P. S. P. V . Kumar, “A comprehensive survey on radio frequency (rf) fingerprinting: Traditional approaches, deep learning, and open challenges,” Computer Networks , vol. 219, p. 109455, Dec. 2022. [Online]. Available: http://dx.doi.org/10.1016/j. comnet.2022.109455

    arXiv 2022

  4. [4]

    A review of radio frequency fingerprinting techniques,

    N. Soltanieh, Y . Norouzi, Y . Yang, and N. C. Karmakar, “A review of radio frequency fingerprinting techniques,” IEEE Journal of Radio Frequency Identification, vol. 4, no. 3, pp. 222–233, 2020

    2020

  5. [5]

    Detection of transient in radio frequency fingerprinting using signal phase,

    J. Hall, M. Barbeau, E. Kranakis et al. , “Detection of transient in radio frequency fingerprinting using signal phase,” in Wireless and optical communications , vol. 9, 2003, p. 13. [Online]. Available: https://api.semanticscholar.org/CorpusID:8404182

    work page 2003

  6. [6]

    Improvements on transient signal detection for rf fingerprinting,

    A. M. Ali, E. Uzundurukan, and A. Kara, “Improvements on transient signal detection for rf fingerprinting,” in 2017 25th Signal Processing and Communications Applications Conference (SIU) . IEEE, May

    work page 2017

  7. [7]

    Direction finding of rogue wi-fi access points using an off-the-shelf mimo–ofdm receiver,

    A. Tzur, O. Amrani, and A. Wool, “Direction finding of rogue wi-fi access points using an off-the-shelf mimo–ofdm receiver,” Physical Communication, vol. 17, p. 149–164, Dec. 2015. [Online]. Available: http://dx.doi.org/10.1016/j.phycom.2015.08.010

    work page doi:10.1016/j.phycom.2015.08.010 2015

  8. [8]

    Wifi-based channel impulse response estimation and localization via multi-band splicing,

    M. B. Khalilsarai, B. Gross, S. Stefanatos, G. Wunder, and G. Caire, “Wifi-based channel impulse response estimation and localization via multi-band splicing,” in GLOBECOM 2020 - 2020 IEEE Global Com- munications Conference, 2020, pp. 1–6

    work page 2020

Show all 32 references
  1. [9]

    Rf fingerprint measurements for the identification of devices in wireless communication networks based on feature reduction and subspace transformation,

    J. L. Padilla, P. Padilla, J. F. Valenzuela-Vald ´es, J. Ram ´ırez, and J. M. G ´orriz, “Rf fingerprint measurements for the identification of devices in wireless communication networks based on feature reduction and subspace transformation,” Measurement, vol. 58, p. 468–475, ...

    2014 doi

  2. [10]

    Transient-based identification of wireless sen- sor nodes,

    B. Danev and S. Capkun, “Transient-based identification of wireless sen- sor nodes,” in 2009 International Conference on Information Processing in Sensor Networks , 2009, pp. 25–36

    2009

  3. [11]

    Specific emitter identification based on hilbert-huang transform-based time-frequency- energy distribution features,

    Y . Yuan, Z. Huang, H. Wu, and X. Wang, “Specific emitter identification based on hilbert-huang transform-based time-frequency- energy distribution features,” IET Commun. , vol. 8, pp. 2404–2412,

  4. [12]

    Rf fingerprint extraction from the energy envelope of an instantaneous transient signal,

    S. Ur Rehman, K. Sowerby, and C. Coghill, “Rf fingerprint extraction from the energy envelope of an instantaneous transient signal,” in 2012 Australian Communications Theory Workshop (AusCTW) . IEEE, Jan. 2012. [Online]. Available: http://dx.doi.org/10.1109/AusCTW.2012. 6164912

    2012 doi

  5. [13]

    Unsupervised drones swarm characterization using rf signals analysis and machine learning methods,

    N. Ashush, S. Greenberg, E. Manor, and Y . Ben-Shimol, “Unsupervised drones swarm characterization using rf signals analysis and machine learning methods,” Sensors, vol. 23, no. 3, p. 1589, Feb. 2023. [Online]. Available: http://dx.doi.org/10.3390/s23031589

    2023 doi

  6. [14]

    Physical layer authentication of internet of things wireless devices through permutation and dispersion entropy,

    G. Baldini, R. Giuliani, G. Steri, and R. Neisse, “Physical layer authentication of internet of things wireless devices through permutation and dispersion entropy,” in 2017 Global Internet of Things Summit (GIoTS) . IEEE, Jun. 2017. [Online]. Available: http://dx.doi.org/10.11...

    2017

  7. [15]

    Transient-based radio frequency fingerprinting with adaptive ensemble of transforms and convolutional neural network,

    G. Baldini, “Transient-based radio frequency fingerprinting with adaptive ensemble of transforms and convolutional neural network,” Electronics Letters , vol. 59, no. 22, Nov. 2023. [Online]. Available: http://dx.doi.org/10.1049/ell2.13032

    2023 doi

  8. [16]

    A radio frequency fingerprinting scheme using learnable signal representation,

    Y . Shao, J. Liu, Y . Zeng, and Y . Gong, “A radio frequency fingerprinting scheme using learnable signal representation,” IEEE Communications Letters, vol. 28, no. 1, p. 73–77, Jan. 2024. [Online]. Available: http://dx.doi.org/10.1109/LCOMM.2023.3336901

    2024

  9. [17]

    Exposing the fingerprint: Dissecting the impact of the wireless channel on radio fingerprinting,

    A. Al-Shawabka, F. Restuccia, S. D’Oro, T. Jian, B. Costa Rendon, N. Soltani, J. Dy, S. Ioannidis, K. Chowdhury, and T. Melodia, “Exposing the fingerprint: Dissecting the impact of the wireless channel on radio fingerprinting,” in IEEE INFOCOM 2020 - IEEE Conference on Compute...

    2020

  10. [18]

    Specific emitter identification via convolutional neural networks,

    L. Ding, S. Wang, F. Wang, and W. Zhang, “Specific emitter identification via convolutional neural networks,” IEEE Communications Letters, vol. 22, no. 12, p. 2591–2594, Dec. 2018. [Online]. Available: http://dx.doi.org/10.1109/LCOMM.2018.2871465

    2018

  11. [19]

    Open set wireless transmitter authorization: Deep learning approaches and dataset considerations,

    S. Hanna, S. Karunaratne, and D. Cabric, “Open set wireless transmitter authorization: Deep learning approaches and dataset considerations,” IEEE Transactions on Cognitive Communications and Networking, vol. 7, no. 1, p. 59–72, Mar. 2021. [Online]. Available: http://dx.doi.org...

    2021

  12. [20]

    The day-after-tomorrow: On the performance of radio fingerprinting over time,

    A. Saeif, S. Savio, and O. Gabriele, “The day-after-tomorrow: On the performance of radio fingerprinting over time,” in Annual Computer Security Applications Conference , ser. ACSAC ’23. ACM, Dec. 2023. [Online]. Available: http://dx.doi.org/10.1145/3627106.3627192

    2023

  13. [21]

    Ai-based rf- fingerprinting framework and implementation using software-defined radios,

    H. Kulhandjian, E. Batz, E. Garcia, S. Vega, S. Velma, M. Kulhandjian, C. D’Amours, B. Kantarci, and T. Mukherjee, “Ai-based rf- fingerprinting framework and implementation using software-defined radios,” in 2023 International Conference on Computing, Networking and Communicat...

    2023

  14. [22]

    Challenges of radio frequency fingerprinting: From data collection to deployment,

    S. Alhazbi, A. Hussain, S. Sciancalepore, G. Oligeri, and P. Papadimitratos, “Challenges of radio frequency fingerprinting: From data collection to deployment,” 2023. [Online]. Available: https://arxiv.org/abs/2310.16406

    2023 arXiv

  15. [23]

    Ruff – rotating uwb for fingerprint,

    T. Ardoin and M. Kholghi, “Ruff – rotating uwb for fingerprint,” 2024. [Online]. Available: https://zenodo.org/doi/10.5281/zenodo.11083153

    2024 doi

  16. [24]

    An image is worth 16x16 words: Transformers for image recognition at scale,

    A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weissenborn, X. Zhai, T. Unterthiner, M. Dehghani, M. Minderer, G. Heigold, S. Gelly, J. Uszkoreit, and N. Houlsby, “An image is worth 16x16 words: Transformers for image recognition at scale,” 2020. [Online]. Available: https://arxi...

    2020 arXiv

  17. [25]

    A survey of transformers,

    T. Lin, Y . Wang, X. Liu, and X. Qiu, “A survey of transformers,” 2021. [Online]. Available: https://arxiv.org/abs/2106.04554

    2021 arXiv

  18. [26]

    Arcface: Additive angular margin loss for deep face recognition,

    J. Deng, J. Guo, N. Xue, and S. Zafeiriou, “Arcface: Additive angular margin loss for deep face recognition,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2019, pp. 4690–4699

    2019

  19. [27]

    Evaluating appearance models for recognition, reacquisition, and tracking,

    D. Gray, S. Brennan, and H. Tao, “Evaluating appearance models for recognition, reacquisition, and tracking,” in Proc. IEEE international workshop on performance evaluation for tracking and surveillance (PETS), vol. 3, no. 5, 2007, pp. 1–7. [Online]. Available: https: //api.se...

    2007

  20. [28]

    Learning mid-level filters for person re-identification,

    R. Zhao, W. Ouyang, and X. Wang, “Learning mid-level filters for person re-identification,” in 2014 IEEE Conference on Computer Vision and Pattern Recognition. IEEE, Jun. 2014. [Online]. Available: http://dx.doi.org/10.1109/CVPR.2014.26

    2014 doi

  21. [29]

    Facenet: A unified embed- ding for face recognition and clustering,

    F. Schroff, D. Kalenichenko, and J. Philbin, “Facenet: A unified embed- ding for face recognition and clustering,” in 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , 2015, pp. 815–823

    2015

  22. [30]

    Deep learning face attributes in the wild,

    Z. Liu, P. Luo, X. Wang, and X. Tang, “Deep learning face attributes in the wild,” in Proceedings of International Conference on Computer Vision (ICCV), December 2015, pp. 3730–3738

    2015

  23. [2014]

    Available: https://api.semanticscholar.org/CorpusID: 23207381

    [Online]. Available: https://api.semanticscholar.org/CorpusID: 23207381

  24. [2017]

    Available: http://dx.doi.org/10.1109/SIU.2017.7960417

    [Online]. Available: http://dx.doi.org/10.1109/SIU.2017.7960417

    2017

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.