Pith. sign in

REVIEW 4 major objections 6 minor 36 references

Towards Secure AI-driven Industrial Metaverse with NFT Digital Twins

T0 review · 4 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read The paper claims that a denoising autoencoder plus a bidirectional GRU can classify digital-twin behavior patterns with 98.30% accuracy, and that dynamic metadata on smart contracts turns this into real-time fake NFT-DT detection.

desk verdict Plausible classifier pilot, but the fake-clone detection claim is untested and the verification algorithm has a blocking bug. read the letter →

arxiv 2412.15716 v1 pith:QVRLCQPX submitted 2024-12-20 cs.CR cs.AIcs.HC

Ravi Prakash , Tony Thomas This is my paper · ORCID
classification cs.CRcs.AIcs.HC
keywords NFTdigitaltwinsindustrialmetaversefakeclonedetectionbehavioralbiometricsdenoisingautoencoderbidirectionalGRUdynamicmetadatasmartcontracts
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 stop counterfeiting of NFT-based digital twins in the industrial metaverse. Existing clone detection compares static metadata, which an attacker can copy, so the paper instead treats each twin's time-series sensor readings as a behavioral fingerprint. A denoising autoencoder compresses a window of 34 readings across 5 features into a 62-element code, and a bidirectional gated recurrent unit (GRU) classifies the code. On data from four heater digital twins, the model reaches 98.30% accuracy. The paper also introduces dynamic metadata whose read-only fields a smart contract updates with the latest encoded behavior, so a verification contract can compare cached and live behavior and declare clones fake.

What carries the argument

The carrying mechanism is the encoded behavioral fingerprint produced by the denoising autoencoder and classified by the bidirectional GRU, together with dynamic metadata whose read-only fields are updated on-chain by the smart contract. The autoencoder serves two purposes at once: it compresses large sensor windows for fast classification and keeps raw behavioral data private by exposing only the 62-element encoding. The verification identity that makes the scheme work is the requirement that a genuine clone and its original share the same live behavioral source, so their metadata updates stay aligned, while a fake clone's metadata diverges from the original's at the next update.

What would settle it

Feed the trained pipeline a fake NFT-DT whose metadata is copied but whose behavioral stream is replayed directly from the genuine twin: Algorithm 2 will encode two identical patterns, return the same class for both, and print 'Verification Successful (genuine clone)'.

Watch

Extended reading notes

Core claim

The paper's claim is that the originality of an NFT digital twin can be verified from behavior rather than from static metadata. The mechanism is a two-stage model: a denoising autoencoder compresses raw behavioral windows into short latent codes, and a bidirectional GRU classifier identifies which of the known digital twins produced each code. Evaluated on 2,640 samples from four heater digital twins, the classifier attains 98.30% accuracy before thresholding and 97.73% after a confidence threshold of 69.50% is applied. In the proposed verification protocol, a smart contract updates read-only metadata with the latest encoded pattern, and Algorithm 2 compares the cached and live predictions; a mismatch declares the NFT-DT a fake clone. The paper argues that because fake clones have behavioral patterns different from the original, the metadata of a fake clone diverges from the original over time, exposing the attack even when the metadata was copied.

Load-bearing premise

A fake clone of a digital twin produces behavioral data that differ from the original's, so a classifier trained only on four genuine heater twins can tell them apart.

Editorial extensions

If this is right

  • Digital twins can be verified in near real time from a compact 62-element behavioral code, so raw sensor readings do not have to leave the owner's control.
  • A confidence threshold around 69.50% lets the system reject behavior from an unseen digital twin instead of forcing it into one of the known classes.
  • Because read-only metadata fields are updated by a smart contract, an attacker cannot manually edit the stored behavioral fingerprint without breaking the verification path.
  • The same DAE-plus-Bi-GRU pipeline can be carried over to NFT avatars by replacing device sensor streams with avatar movement data.
  • Fake clones created by copying a URI or metadata are exposed as soon as their live behavior stops matching the original's dynamic metadata.

Reading between the lines

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

  • The end-to-end fake-clone result is an extrapolation from the paper's own experiments: the classifier is evaluated on four genuine heater digital twins, and the metadata-copy attack is demonstrated only with similarity scores, not by running a fake clone through Algorithm 2.
  • Because Algorithm 2 compares encodings of cached and live behavior, a clone that replays the original's exact data stream would pass the verification; closing that gap requires an independent source of trust for the live stream, which the paper does not specify.
  • The paper's Section VII explicitly flags the lack of out-of-class data and the deployment difficulty of a multiclass model, so the natural next test is to train on three digital twins and treat the fourth as unseen to measure how often the threshold rule rejects it.
  • The security of dynamic metadata rests on the smart contract being truly tamper-proof, since the whole argument fails if the owner of a fake NFT-DT can update the read-only fields.
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.

Referee Report

4 major / 6 minor

Summary. The paper proposes a hybrid denoising-autoencoder and bidirectional-GRU architecture to classify behavioural patterns of NFT digital twins, together with a smart-contract-based 'dynamic metadata' mechanism for authenticity verification in an industrial metaverse. The authors evaluate the classifier on a public Kaggle dataset of four heater digital twins, encoding 34x5 samples into 62-element vectors and reporting 98.30% closed-set accuracy (Table VI) and 97.73% accuracy after applying a confidence threshold tau (Table VII). They also describe a Ganache/Hardhat testbed for minting and cloning NFT-DTs and report metadata-similarity scores for cloned NFTs (Table IV). The central security claim is that this system can detect fake NFT-DTs in real time; this claim is not supported by the current experiments.

Significance. If the fake-NFT-DT detection claim were established, the paper would make a useful contribution by combining behavioural biometrics-style analysis with NFT ownership for digital twins, and by highlighting the insufficiency of static metadata. The authors deserve credit for building a concrete blockchain testbed and for identifying a real problem in the industrial-metaverse literature. However, the evaluation never includes a fake clone or out-of-class pattern; the reported accuracy is a closed-set identification result among four genuine DTs. The security claim therefore rests on an unvalidated assumption (Table II) that fake clones exhibit dissimilar behavioural patterns, and on a threshold that is tuned on the test set. As it stands, the paper demonstrates a four-class classifier and a metadata-update concept, not a fake-NFT-DT detector.

major comments (4)
  1. [Section VI.B, Tables VI and VII] The evaluation contains only 528 genuine test samples from the four enrolled heaters; no fake NFT-DT sample, replayed stream, or perturbed behavioural pattern is included. The reported 98.30% and 97.73% accuracies therefore measure closed-set identity recognition among genuine DTs, not the ability to detect a counterfeit NFT-DT. Table II's premise that a fake clone has 'Dissimilar behavioural patterns' is asserted but never tested; an attacker who copies both the metadata and the behavioural data stream would produce identical patterns and would pass the equality check in Algorithm 2. The central detection claim is unsupported by the experiments.
  2. [Section VI.B, Fig. 7 and Table VII] The optimal threshold tau = 69.50% is selected by 'analysis of true and misclassified samples' on the test set, and the final accuracy, FRR, and FP/FN counts in Table VII are computed on the same 528 samples. This is test-set fitting and does not provide an unbiased estimate of performance on unseen DTs or unseen attacks. The threshold should be chosen on a validation set (or via nested cross-validation) and reported with confidence intervals.
  3. [Section V.D, Algorithm 2] Algorithm 2 contains an undefined condition 'N is odd' on line 12; N is not declared in the algorithm, and if the condition is false the verification function produces no output. More importantly, the verification logic compares only the predicted classes of the cached and latest patterns, so it cannot distinguish a genuine DT from a replay of a recorded genuine stream. The dynamic-metadata argument in Eqs. (6)-(7) assumes that at time t' the fake clone's behaviour diverges from the original, but this is the same unvalidated assumption from Table II and is not demonstrated by any experiment.
  4. [Section V.C and Table III] The completeness scores in Table III are computed on genuine samples only, so they measure per-class recall, not resistance to adversarial spoofing. The statement that an attacker 'could only succeed 6 (1.14%) times to spoof the target' conflates natural misclassification among enrolled DTs with a deliberate spoofing attack. No attacker-controlled inputs are evaluated, so Table III does not support the robustness claim.
minor comments (6)
  1. [Section III.B, Eq. (4)] The sentence 'according to the Eqn. 3, Eqn. 4, and Eqn. 4' should refer to Eqs. (3), (4), and (5); the current reference repeats Eq. (4).
  2. [Sections III.A and V.D] There are several typos, including 'dveloped' (Section III.A), 'temper-proof' (Section V.D), and 'signifient' (Section V.D); these should be corrected.
  3. [Section VI.B] The phrase '62 character' should read '62-element vector'; the encoded representation is an array of numbers, not a character string.
  4. [Table VII] The entry TN=3 is unexplained, since the test set contains no out-of-class samples; please define what a true negative is in this closed-set evaluation.
  5. [Section VII] The conclusion admits that 'sufficient OOC data' would be needed for more realistic results; this limitation should be acknowledged in the abstract and in the results section, because it directly affects the headline accuracy and the fake-detection claim.
  6. [Section VI.B] The paper does not compare the proposed DAE+Bi-GRU pipeline against simpler baselines (e.g., classification on raw patterns or a one-class anomaly detector); such a comparison would help assess the benefit of the encoding step.

Circularity Check

2 steps flagged · score 6.0 of 10

OOC threshold is fitted to the test set; fake-clone detection is defined rather than demonstrated.

  1. fitted input called prediction [Section VI.B (Fig. 7, Table VII)]
    "By analyzing the confidence levels associated with true and false predictions, we observed a clear discrepancy. Specifically, true predictions relatively exhibited higher confidence scores compared to false ones. Therefore, we introduced a threshold parameter, τ, to effectively detect OOC DT instances. ... Through the analysis of true and misclassified samples, we determined the optimal threshold value to be τ = 69.50%."

    The threshold τ is chosen by inspecting the true and misclassified labels of the same 528 test samples that are then used to report final performance. Table VII recomputes accuracy (97.73%), FRR, TP, FP, and FN on that identical test set after applying the fitted τ. The OOC-rejection performance is therefore a post-hoc fit to the test labels, not an independent prediction; the reported numbers are forced by the fitting procedure by construction.

  2. self definitional [Table II and Section V.D (Eq. 6-7, Algorithm 2)]
    "Table II: 'Fake Clone: Dissimilar behavioural patterns'; Section V.D: 'The change occurs due to the difference in behavioural pattern of genuine and fake clones at time t′.' (Eq. 7: ∆M αi ft′ ≠ ∆M αi ot′)."

    The fake-detection claim rests on defining a fake clone as having 'Dissimilar behavioural patterns' (Table II) and then treating Algorithm 2's comparison Pα ≠ Pc as the detector. No experiment in Section VI.B includes a fake clone or replayed stream; only four genuine heater DTs are classified. Thus 'detection of fake NFT-DTs' reduces to the definitional assumption that fakes behave differently. A clone that copies both metadata and the live data stream yields Pα = Pc and is declared 'Verification Successful' by construction, so the security conclusion is an artifact of the definition rather than an empirically tested capability.

full rationale

The paper's core classifier (DAE + Bi-GRU) achieves 98.30% accuracy on held-out test samples of four known heater DTs (Table VI), which is a legitimate multi-class classification result and not circular. However, two evaluation steps compromise the security claims. First, the OOC threshold τ is explicitly selected by analyzing true and misclassified samples on the test set (Fig. 7), and the final accuracy/FRR are recomputed on that same test set (Table VII); this is a fitted-input-called-prediction circularity for the OOC detection capability. Second, the fake-clone detection claim is self-definitional: Table II defines a fake clone as having 'Dissimilar behavioural patterns', Algorithm 2 declares fake iff Pα ≠ Pc, and no experiment tests a fake clone, a replayed stream, or any genuinely anomalous pattern. The spoof-resistance claim in Section V.C is derived from confusion among genuine DTs only, so it does not independently validate fake-clone detection. Additionally, Algorithm 2 contains an undefined condition 'N is odd', leaving the verification logic incomplete as written. Overall, the underlying classifier is non-circular, but the security predictions around τ and fake-clone detection are partially forced by construction, yielding a partial circularity score of 6.

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

The central claims rest on the assumption that fake clones have distinguishable behavior, on the chosen neural architectures (with k'=62 and the tau threshold as fitted values), and on the correctness of the proposed verification algorithm. The algorithm's logic bug and the untested assumptions make the ledger heavy for a security claim.

free parameters (3)
  • tau = 69.50%
    Rejection threshold for out-of-class detection; chosen post hoc on the test set, so final metrics are in-sample.
  • encoding size k' = 62
    Output dimension of the autoencoder, chosen during hyperparameter tuning; no principled derivation provided.
  • encoder model epsilon_i = trained per NFT-DT
    Each NFT-DT deploys its own encoder; parameters and training details are not fully specified.
assumptions (3)
  • domain assumption A fake clone of an NFT-DT will exhibit behavioral patterns different from the original.
    Stated in Table II and used as the basis for pattern-based clone detection; never experimentally verified.
  • domain assumption Smart contracts are tamper-proof and the encoder model address is trusted.
    The dynamic metadata security relies on the smart contract being unable to be modified by attackers, as assumed in Section V.D.
  • ad hoc to paper The verification flow's condition 'N is odd' is meaningful.
    Algorithm 2 prints 'Verification Successful' only if 'N is odd', with N never defined; the algorithm's correctness depends on this unexplained condition.
invented entities (1)
  • dynamic metadata (Delta M)
    purpose: Stores encoded behavioral patterns in NFT metadata for automatic authenticity verification via smart contracts.
    Proposed as a concept; no implementation or security proof is given beyond a high-level algorithm sketch.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Towards Secure AI-driven Industrial Metaverse with NFT Digital Twins." pith.science (2026). https://pith.science/paper/QVRLCQPX

@misc{pith2026241215716,
  author       = {Pith},
  title        = {Pith review of: Towards Secure AI-driven Industrial Metaverse with NFT Digital Twins},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QVRLCQPX}},
  note         = {Machine review of arXiv:2412.15716}
}
read the original abstract

The rise of the industrial metaverse has brought digital twins (DTs) to the forefront. Blockchain-powered non-fungible tokens (NFTs) offer a decentralized approach to creating and owning these cloneable DTs. However, the potential for unauthorized duplication, or counterfeiting, poses a significant threat to the security of NFT-DTs. Existing NFT clone detection methods often rely on static information like metadata and images, which can be easily manipulated. To address these limitations, we propose a novel deep-learning-based solution as a combination of an autoencoder and RNN-based classifier. This solution enables real-time pattern recognition to detect fake NFT-DTs. Additionally, we introduce the concept of dynamic metadata, providing a more reliable way to verify authenticity through AI-integrated smart contracts. By effectively identifying counterfeit DTs, our system contributes to strengthening the security of NFT-based assets in the metaverse.

Figures

Figures reproduced from arXiv: 2412.15716 by the authors.

Figure 1
Figure 1. Components of Industrial Metaverse A. Pattern Encoding Autoencoders are artificial neural networks (ANNs) designed to compress and reconstruct input data. These models consist of an encoder component that compresses the input into a latent representation as Pe and a decoder that reconstructs the original input from this latent representation or encoding. The sample of a raw behavioural pattern of a DT contains the t… view at source ↗
Figure 2
Figure 2. DT classification from raw behavioural patterns [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Data flow in a Bi-GRU [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (4 more)
Figure 5
Figure 5. Figure 5: Features of the test bed A. Experimental Set-up for DT Behavioural Pattern Recogni￾tion We consider the scenario of the industrial metaverse where the digital twin of heaters is to be used as NFT-DTs. As discussed in Section III, the proposed method of NFT comparison (…
Figure 6
Figure 6. Figure 6: However, one major problem with such a multiclass classification model is that it can not accurately identify the pattern from an OOC/unseen DT. Hence, the model currently [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 6
Figure 6. Figure 6: Soundness and Completeness score of proposed solution for 4 NFT [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: Optimal threshold τ was found to be 69.50% The final results obtained after setting the τ are provided in Table VII. Here false positive (FP) represents the misclassified samples with a confidence higher than τ and false negative (FN) represents the count of correct cl…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 34 canonical work pages

  1. [1]

    Industrial metaverse: Connotation, features, technologies, applications and challenges,

    Z. Zheng, T. Li, B. Li, X. Chai, W. Song, N. Chen, Y . Zhou, Y . Lin, and R. Li, “Industrial metaverse: Connotation, features, technologies, applications and challenges,” in Asian Simulation Conference. Springer, 2022, pp. 239–263

    work page 2022

  2. [2]

    Digital twins for building industrial meta- verse,

    Z. Lyu and M. Fridenfalk, “Digital twins for building industrial meta- verse,” Journal of Advanced Research , 2023

    work page 2023

  3. [3]

    Exploring the intersection of metaverse, digital twins, and artificial intelligence in training and maintenance,

    M. Bordegoni and F. Ferrise, “Exploring the intersection of metaverse, digital twins, and artificial intelligence in training and maintenance,” Journal of Computing and Information Science in Engineering , vol. 23, no. 6, p. 060806, 2023

    work page 2023

  4. [4]

    Build a secure smart city by using blockchain and digital twin,

    Y . Song and S. Hong, “Build a secure smart city by using blockchain and digital twin,” International Journal of Advanced Science and Con- vergence, vol. 3, no. 3, pp. 9–13, 2021

    work page 2021

  5. [5]

    A dynamic hierarchical framework for iot-assisted digital twin synchronization in the metaverse,

    Y . Han, D. Niyato, C. Leung, D. I. Kim, K. Zhu, S. Feng, X. Shen, and C. Miao, “A dynamic hierarchical framework for iot-assisted digital twin synchronization in the metaverse,” IEEE Internet of Things Journal, vol. 10, no. 1, pp. 268–284, 2022

    work page 2022

  6. [6]

    Blockchain technology for cybersecurity: A text mining literature analysis,

    R. Prakash, V . Anoop, and S. Asharaf, “Blockchain technology for cybersecurity: A text mining literature analysis,” International Journal of Information Management Data Insights , vol. 2, no. 2, p. 100112, 2022

    work page 2022

  7. [7]

    How the rarity influences aesthetic experience of 3d profile picture nfts,

    T. Jing and L. Wang, “How the rarity influences aesthetic experience of 3d profile picture nfts,” in Proceedings of the Tenth International Symposium of Chinese CHI , 2022, pp. 151–160

    work page 2022

  8. [8]

    Nft culture: A new era,

    M. Gupta, D. Gupta, and A. Duggal, “Nft culture: A new era,” Scientific Journal of Metaverse and Blockchain Technologies , vol. 1, no. 1, pp. 57–62, 2023

    work page 2023

Show all 36 references
  1. [9]

    A privacy awareness framework for nft avatars in the metaverse,

    D. Zelenyanszki, Z. H ´ou, K. Biswas, and V . Muthukkumarasamy, “A privacy awareness framework for nft avatars in the metaverse,” in 2023 International Conference on Computing, Networking and Communica- tions (ICNC). IEEE, 2023, pp. 431–435

    2023

  2. [10]

    Behavioural biometrics in VR: Identifying people from body motion and relations in virtual reality,

    K. Pfeuffer, M. J. Geiger, S. Prange, L. Mecke, D. Buschek, and F. Alt, “Behavioural biometrics in VR: Identifying people from body motion and relations in virtual reality,” in Proceedings of the 2019 CHI Conference on Human Factors in Computing Systems , 2019, pp. 1–12

    2019

  3. [11]

    Clustering avatars behavioursfrom virtual worlds interactions,

    G. Orgaz et al. , “Clustering avatars behavioursfrom virtual worlds interactions,” in International Workshop On Web Intelligence & Communities-WI&C, vol. 12, no. 4, 2012, pp. 1–7

    2012

  4. [12]

    A metaverse: Taxonomy, components, applications, and open challenges,

    S.-M. Park and Y .-G. Kim, “A metaverse: Taxonomy, components, applications, and open challenges,” IEEE access , vol. 10, pp. 4209– 4251, 2022

    2022

  5. [13]

    Applications of the metaverse in medicine and healthcare,

    T. Hulsen, “Applications of the metaverse in medicine and healthcare,” Advances in Laboratory Medicine/Avances en Medicina de Laboratorio , vol. 5, no. 2, pp. 159–165, 2024

    2024

  6. [14]

    Counseling psychological understanding and considerations of the metaverse: A theoretical review,

    K. Lee, “Counseling psychological understanding and considerations of the metaverse: A theoretical review,” in Healthcare, vol. 11, no. 18. MDPI, 2023, p. 2490

    2023

  7. [15]

    The rising trend of metaverse in education: Challenges, opportunities, and ethical considerations,

    S. Kaddoura and F. Al Husseiny, “The rising trend of metaverse in education: Challenges, opportunities, and ethical considerations,” PeerJ Computer Science, vol. 9, p. e1252, 2023

    2023

  8. [16]

    Metaverse and education: A panoramic review,

    C. O. Lepez, “Metaverse and education: A panoramic review,” Metaverse Basic and Applied Research , vol. 1, pp. 2–2, 2022

    2022

  9. [17]

    Metaverse application in power systems,

    B. Zhou, C. Xi, G. Li, and B. Yang, “Metaverse application in power systems,” Power Generation Technology, vol. 43, no. 1, pp. 1–9, 2022

    2022

  10. [18]

    Metaverse and social view,

    S. Cheng, “Metaverse and social view,” in Metaverse: Concept, Content and Context. Springer, 2023, pp. 107–122

    2023

  11. [19]

    The social metaverse: Battle for privacy,

    B. Falchuk, S. Loeb, and R. Neff, “The social metaverse: Battle for privacy,” IEEE technology and society magazine , vol. 37, no. 2, pp. 52–61, 2018

    2018

  12. [20]

    Metaomnicity: Toward immersive urban metaverse cy- berspaces using smart city digital twins,

    K. Kuru, “Metaomnicity: Toward immersive urban metaverse cy- berspaces using smart city digital twins,” IEEE Access , vol. 11, pp. 43 844–43 868, 2023

    2023

  13. [21]

    Security risk assessment of metaverse based healthcare systems based on common vulnerabilities and exposures (cve),

    R. Prakash, G. R. Nayar, and T. Thomas, “Security risk assessment of metaverse based healthcare systems based on common vulnerabilities and exposures (cve),” in 2023 IEEE International Conference on Recent Advances in Systems Science and Engineering (RASSE) . IEEE, 2023, pp. 1–10

    2023

  14. [22]

    Metaverse-driven remote management solution for scene-based energy storage power stations,

    Y . Deng, Z. Weng, and T. Zhang, “Metaverse-driven remote management solution for scene-based energy storage power stations,” Evolutionary Intelligence, vol. 16, no. 5, pp. 1521–1532, 2023

    2023

  15. [23]

    Nfts in healthcare: vision, opportunities, and challenges,

    A. Musamih, K. Salah, R. Jayaraman, I. Yaqoob, D. Puthal, and S. Ellahham, “Nfts in healthcare: vision, opportunities, and challenges,” IEEE consumer electronics magazine , vol. 12, no. 4, pp. 21–32, 2022

    2022

  16. [24]

    Nft digital twins: a digitalization strategy to preserve and sustain miao silver craftsmanship in the metaverse era,

    M. Wang and N. Lau, “Nft digital twins: a digitalization strategy to preserve and sustain miao silver craftsmanship in the metaverse era,” Heritage, vol. 6, no. 2, pp. 1921–1941, 2023

    1921

  17. [25]

    Unique identification of 50,000+ virtual reality users from head & hand motion data,

    V . Nair, W. Guo, J. Mattern, R. Wang, J. F. O’Brien, L. Rosenberg, and D. Song, “Unique identification of 50,000+ virtual reality users from head & hand motion data,” in 32nd USENIX Security Symposium (USENIX Security 23) , 2023, pp. 895–910

    2023

  18. [26]

    Deep learning for source camera identification on mobile devices,

    D. Freire-Obreg ´on, F. Narducci, S. Barra, and M. Castrillon-Santana, “Deep learning for source camera identification on mobile devices,” Pattern Recognition Letters, vol. 126, pp. 86–91, 2019

    2019

  19. [27]

    A compar- ative study of pattern recognition techniques for quality evaluation of telecommunications software,

    T. M. Khoshgoftaar, D. L. Lanning, and A. S. Pandya, “A compar- ative study of pattern recognition techniques for quality evaluation of telecommunications software,” IEEE Journal on Selected Areas in Communications, vol. 12, no. 2, pp. 279–291, 1994

    1994

  20. [28]

    An approach for malware behavior identification and classification,

    M. F. Zolkipli and A. Jantan, “An approach for malware behavior identification and classification,” in 2011 3rd international conference on computer research and development, vol. 1. IEEE, 2011, pp. 191–194

    2011

  21. [29]

    Digital twin-based intrusion detection for industrial con- trol systems,

    S. A. Varghese, A. D. Ghadim, A. Balador, Z. Alimadadi, and P. Pa- padimitratos, “Digital twin-based intrusion detection for industrial con- trol systems,” in 2022 IEEE International Conference on Pervasive Computing and Communications Workshops and other Affiliated Events (Pe...

    2022

  22. [30]

    Convolutional rnn: an enhanced model for extracting features from sequential data,

    G. Keren and B. Schuller, “Convolutional rnn: an enhanced model for extracting features from sequential data,” in 2016 International Joint Conference on Neural Networks (IJCNN). IEEE, 2016, pp. 3412–3419

    2016

  23. [31]

    A recurrent latent variable model for sequential data,

    J. Chung, K. Kastner, L. Dinh, K. Goel, A. C. Courville, and Y . Bengio, “A recurrent latent variable model for sequential data,” Advances in neural information processing systems , vol. 28, 2015

    2015

  24. [32]

    The vanishing gradient problem during learning recurrent neural nets and problem solutions,

    S. Hochreiter, “The vanishing gradient problem during learning recurrent neural nets and problem solutions,” International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems , vol. 6, no. 02, pp. 107–116, 1998

    1998

  25. [33]

    Empirical evaluation of gated recurrent neural networks on sequence modeling,

    J. Chung, “Empirical evaluation of gated recurrent neural networks on sequence modeling,” arXiv preprint arXiv:1412.3555 , 2014

    2014 arXiv

  26. [34]

    Gate-variants of gated recurrent unit (gru) neural networks,

    R. Dey and F. M. Salem, “Gate-variants of gated recurrent unit (gru) neural networks,” in 2017 IEEE 60th international midwest symposium on circuits and systems (MWSCAS) . IEEE, 2017, pp. 1597–1600

    2017

  27. [35]

    Digital twin gadget health,

    P. Akella, “Digital twin gadget health,” https://www.kaggle.com/datasets/ prasannaakella/digital-twin-gadget-health, 2021

    2021

  28. [36]

    Cloning and teleporting avatars across workstations and mobile devices in collaborative virtual environments: Clipboard operations for virtual reality,

    O. N. N. Fernando, G. Saito, U. Duminduwardena, Y . Tanno, and M. Co- hen, “Cloning and teleporting avatars across workstations and mobile devices in collaborative virtual environments: Clipboard operations for virtual reality,” Proc. ICIA, vol. 5, 2005

    2005

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.