Pith. sign in

REVIEW 5 major objections 7 minor 34 references

FedMultiEmo: Real-Time Emotion Recognition via Multimodal Federated Learning

T0 review · 5 major / 7 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read FedMultiEmo claims that decision-level fusion of a federated facial-expression CNN and a federated physiological random forest reaches 87% emotion-recognition accuracy—matching the centralized baseline while keeping all raw data local.

desk verdict The prototype is real and the integration story is plausible, but the headline 87% fusion accuracy doesn't follow from the stated majority-vote rule and the reported numbers are inconsistent. read the letter →

arxiv 2507.15470 v2 pith:6BHJKEMA submitted 2025-07-21 cs.LG

Baran Can Gül , Suraksha Nadig , Stefanos Tziampazis , Nasser Jazdi , Michael Weyrich This is my paper · ORCID
classification cs.LG
keywords federatedlearningemotionrecognitionmultimodalfusiondrivermonitoringedgeAIprivacyrandomforestconvolutionalneuralnetwork
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 emotion recognition in a car can be federated—trained across devices without raw data leaving them—without giving up accuracy. It pairs a convolutional neural network that reads facial expressions with a random forest that reads heart rate, electrodermal activity, and skin temperature, and fuses the two at the decision level by majority vote. On a public facial-expression dataset plus a lab-collected physiological dataset, the fused model reaches 87% accuracy, while the vision-only and physiology-only models reach 77% and 74%, respectively, matching the centralized baseline. If the result holds, adaptive driver-assistance and comfort features can be trained on sensitive occupant data while that data stays inside the vehicle.

What carries the argument

The load-bearing mechanism is decision-level majority-vote fusion of two heterogeneous classifiers trained under federated averaging. Each client trains a small CNN on its own facial images and a 200-tree random forest on hand-crafted physiological features (heart-rate variability, maximum electrodermal activity, and temperature fluctuation over five-second windows); only the model updates are sent to the server, and the server aggregates them by a FedAvg rule weighted by each client's local data volume. At inference, each modality produces a vote over the seven emotion classes, and Eq. (16) selects the class with the highest summed indicator votes. Because fusion happens after prediction, models of completely different types—a gradient-trained deep network and a non-gradient ensemble—can be combined without sharing gradients, features, or raw samples.

What would settle it

Count the number of test samples on which the two modalities disagree and re-run the fusion evaluation with a fixed tie-break (for example, always choose the visual prediction). If ties are frequent and the accuracy moves by more than a few points when the tie-break changes, the 87% figure is not robust; likewise, subject-level cross-validation of the physiological dataset would settle whether the 74% and the fusion gain generalize.

Watch

Extended reading notes

Core claim

FedMultiEmo claims that multimodal, decision-level fusion lets a federated system match or exceed centralized accuracy for real-time emotion recognition. The central discovery is concrete: a federated CNN on facial images reaches about 77% accuracy, a federated random forest on physiological features reaches 74%, and their majority-vote fusion reaches 87%—a 10-point gain over the vision-only model—while only model updates, not raw images or signals, are exchanged. The fusion works because the two modalities fail in complementary ways: vision degrades under poor lighting or occlusion, while physiological signals are unaffected by the camera but vary across individuals. The prototype also shows that the approach is deployable: it converges in about 18 rounds with an average round time of 120 seconds and a per-client memory footprint under 200 MB on low-cost edge hardware.

Load-bearing premise

The 87% fusion result rests on a self-collected physiological dataset whose number of subjects, per-class sample counts, and label-validation procedure are not reported, and on an unspecified tie-breaking rule for the majority vote.

Editorial extensions

If this is right

  • Adaptive in-vehicle responses such as climate, seat, and stress-relief actions could be triggered by emotion state without transmitting facial images or physiological streams to a central server.
  • The roughly 10-point accuracy gain from fusion over the visual-only model indicates that the system can keep working when the camera view degrades, because the physiological channel still votes.
  • Because training converges in about 18 rounds with sub-200 MB per-client memory, the approach fits the compute and communication budgets of low-cost in-car edge devices rather than requiring cloud GPUs.
  • The same majority-vote design can be reused for other heterogeneous model pairs, since only output class votes are exchanged at fusion time.

Reading between the lines

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

  • Subject-level rather than sample-level evaluation of the self-collected physiological dataset would test whether the 74% random-forest figure and the fusion gain persist across drivers, since physiological responses vary strongly between individuals.
  • A confidence-weighted fusion rule, which the paper lists as future work, could push accuracy beyond 87% or reveal which emotion classes carry the fusion gain; the confusion matrices suggest disgust benefits most from adding the physiological modality.
  • The same decision-level design could be adapted to other vehicle-sensing pairs—for example, voice plus facial expression—whenever the modalities are heterogeneous enough that a shared neural architecture would be wasteful.
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

5 major / 7 minor

Summary. The paper introduces FedMultiEmo, a privacy-preserving multimodal federated learning framework for real-time in-vehicle emotion recognition. Visual features from facial images are classified by a CNN, physiological signals (heart rate, electrodermal activity, skin temperature) are classified by a Random Forest, and the two modality predictions are combined by decision-level majority voting. Training uses Federated Averaging across three Raspberry Pi clients orchestrated by a Flower server, with client updates weighted by local dataset size. The paper reports 77% accuracy for the federated CNN, 74% for the Random Forest, and 87% for the fused system on FER2013 plus a self-collected physiological dataset, along with convergence in 18 rounds, an average round time of 120 seconds, and a per-client memory footprint below 200 MB. The claimed contribution is a practical, privacy-preserving multimodal emotion recognition system that matches centralized baselines while keeping raw data local.

Significance. If the reported results are correct, FedMultiEmo would be a useful systems demonstration: it combines two complementary modalities in a federated setting, provides an end-to-end edge-cloud deployment on Raspberry Pi clients, uses the public FER2013/FERPlus benchmark for the visual modality, and reports confusion matrices for all three classification configurations. The explicit comparison of individual, centralized, and federated training costs is also informative. However, the central numerical claims rest on three load-bearing points that are not yet established: the fusion rule as written cannot produce the reported 87% accuracy, the self-collected physiological dataset is not characterized enough for reproducibility, and several reported experimental constants are internally inconsistent. The paper does not provide code, data, or machine-checked proofs, so the credibility of the headline numbers depends entirely on the completeness and correctness of the experimental description. These issues are fixable in a revision, but they currently prevent the reader from verifying the central claim.

major comments (5)
  1. [Section III-F, Eq. (16)] The stated majority-vote fusion rule is undefined whenever the CNN and Random Forest predictions disagree, and no tie-breaking rule is given. With two classifiers and seven classes, any disagreement produces two classes tied at one vote, so the argmax in Eq. (16) is not well defined. If ties were always resolved to the visual prediction, the fused accuracy would equal the CNN accuracy (77% or 78%); if always to the physiological prediction, it would equal the Random Forest accuracy (74%); if broken uniformly at random, the expected fused accuracy would be roughly the average of the two, about 75.5%. The reported 87% fused accuracy therefore cannot follow from Eq. (16) as written. The paper must either state the actual tie-breaking rule and show how it yields 87%, or correct Eq. (16) to describe the actually implemented fusion (e.g., soft voting over probability vectors) and report the corresponding accuracy.
  2. [Section IV-B] The bespoke physiological dataset is the sole basis for the 74% Random Forest accuracy and the 87% fusion accuracy, but it is not characterized sufficiently for reproducibility. The paper does not state the number of subjects, the number of sessions, the per-class sample counts, how the video-elicited ground-truth labels were validated (e.g., self-report, multiple annotators, or external measures), or how train/test splits were constructed. It is also unclear how the CNN trained on FER2013 and the Random Forest trained on this custom set are evaluated jointly at test time, since no paired test set or data synchronization procedure is described. Without this information, the fusion result cannot be interpreted or reproduced.
  3. [Section IV-D and Table I] The number of federated rounds and the per-round time are reported inconsistently: the abstract and conclusion state 18 rounds, Section IV-D states 20 global rounds, and Table I lists 10 rounds; the abstract gives an average round time of 120 seconds while Table I gives 430 seconds per round. These contradictions directly affect the convergence and real-time claims and must be reconciled with a single consistent measurement protocol.
  4. [Section III-G and Section IV-D] The federated aggregation of the Random Forest is not specified. Equation (17) is a FedAvg update for weight vectors, but a 200-tree Random Forest does not have a natural weight-average representation, and the text does not describe how the server aggregates tree ensembles across the three clients. Without this detail, the claim that the physiological model is trained in a federated manner is incomplete.
  5. [Section IV-E.1] The reported CNN accuracy is 77% in the abstract but 78% in Section IV-E.1, and the precision/recall values for the Random Forest (67%/73%) and CNN (77%/77%) are not reconciled with the accuracy figures for the same models. Since these numbers appear in the paper's central performance summary, they need to be made consistent and traceable to specific evaluation runs.
minor comments (7)
  1. [Section III-B, Eq. (4)] Equation (4) states that a 4th-order filter has T = 4 coefficients, but a 4th-order Butterworth filter has more than four coefficients; the notation conflates filter order with filter length and should be corrected.
  2. [Section III-B, Eq. (5)] The moving average in Eq. (5) is written with s[t - i] rather than the filtered signal from Eq. (4); the notation should be made consistent so that the smoothing step is clear.
  3. [Section III-C, Eq. (8)] Equation (8) defines HRV as the mean absolute successive difference of heart rate samples, which is not the standard definition of heart rate variability; additionally, with 1 Hz sampling and a 5-second window, the feature is computed over very few samples and its statistical reliability should be discussed.
  4. [References [11] and [26]] References [11] and [26] both appear to describe the same Fed-PhyERS system but with different authors and venues; these should be unified or clearly distinguished.
  5. [Section III-G] The phrase 'personalized Federated Averaging' is used, but Eq. (17) is standard FedAvg with client weighting by dataset size; the paper should clarify what personalization mechanism beyond standard weighting is being claimed.
  6. [Figure 2] The confusion matrices in Fig. 2 are difficult to read at print resolution; the class labels, color scales, and percentage values should be enlarged or a higher-resolution version provided.
  7. [Section III-A] The number of clients N in Eqs. (1) and (17) is never stated explicitly in the problem formulation; Section IV-A later says three clients, but the notation would benefit from an explicit statement of N in the setup.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the accuracy claims are measured outcomes and self-citations are background; the Eq. 16 tie-break inconsistency is a correctness issue, not a circular reduction.

full rationale

The central claim (77%/74%/87% accuracy) is reported from evaluation on FER2013 and a self-collected physiological set; it is not obtained by solving the model's equations, so it cannot reduce by construction to the model's fitted parameters. Eq. 16 defines hard majority voting over two classifiers; with seven classes and any disagreement both candidate classes receive one vote, so the argmax is undefined. The reported 87% therefore does not follow from Eq. 16 as written, but that is an internal-consistency/reporting defect, not a circular derivation: no quantity in Eq. 16 is fitted so as to force the fusion accuracy, and the individual accuracies are externally benchmarked (FER2013 with FERPlus relabeling). Self-citations [2], [3], [7], [8], [32]-[34] appear only in related work or for NTP time-base support; none is cited as a uniqueness theorem or as justification for an otherwise unmotivated ansatz. The FedAvg aggregation in Eq. 17 is standard from [25] and weighted by local sample counts; it is not defined in terms of the target accuracy. No renamed known result is presented. Thus no circular step is identifiable; the paper's main risks are missing dataset characterization and an undefined tie-break, which belong to correctness/reproducibility review, not circularity analysis.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

No new physical entities are introduced. The central results depend on hand-chosen hyperparameters (window size, client count, local epochs, tree count) and on unstated assumptions about label validity and fusion tie-breaking. The FedAvg aggregation is standard from the cited McMahan et al. work.

free parameters (4)
  • Physiological feature window size = 5 s
    Chosen by hand in Section III.C.b; defines HRV, EDA max, and temperature range features; no ablation is reported.
  • Number of FL clients = 3
    Defines FedAvg aggregation in Eq. 17 and all measured round times; no scaling study to larger fleets.
  • Local epochs per round E = 4
    Set in Section III.D; the convergence claim of 18 rounds depends on this choice.
  • Random Forest tree count = 200
    Set in Section IV.C.b; standard but hand-chosen, affects physiological classifier performance.
assumptions (3)
  • domain assumption Emotion labels elicited by video clips are accurate ground truth for all subjects
    The custom physiological dataset labels are synchronized to the video timeline (Section IV.B.b), but no inter-rater agreement or physiological validation is reported.
  • domain assumption FER2013/FERPlus labels transfer to in-vehicle driver faces
    FER2013 is a general web-sourced facial expression dataset (Section IV.B.a); the paper applies it directly to the automotive setting without evaluating the domain shift.
  • ad hoc to paper Majority voting between two classifiers is a well-defined fusion rule
    Eq. 16 sums two indicator functions; with exactly two modalities, ties occur whenever the two classifiers disagree, and no tie-break is specified.

how reviews work

0 comments
Cite this review

Pith. "Pith review of FedMultiEmo: Real-Time Emotion Recognition via Multimodal Federated Learning." pith.science (2026). https://pith.science/paper/6BHJKEMA

@misc{pith2026250715470,
  author       = {Pith},
  title        = {Pith review of: FedMultiEmo: Real-Time Emotion Recognition via Multimodal Federated Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6BHJKEMA}},
  note         = {Machine review of arXiv:2507.15470}
}
read the original abstract

In-vehicle emotion recognition underpins adaptive driver-assistance systems and, ultimately, occupant safety. However, practical deployment is hindered by (i) modality fragility - poor lighting and occlusions degrade vision-based methods; (ii) physiological variability - heart-rate and skin-conductance patterns differ across individuals; and (iii) privacy risk - centralized training requires transmission of sensitive data. To address these challenges, we present FedMultiEmo, a privacy-preserving framework that fuses two complementary modalities at the decision level: visual features extracted by a Convolutional Neural Network from facial images, and physiological cues (heart rate, electrodermal activity, and skin temperature) classified by a Random Forest. FedMultiEmo builds on three key elements: (1) a multimodal federated learning pipeline with majority-vote fusion, (2) an end-to-end edge-to-cloud prototype on Raspberry Pi clients and a Flower server, and (3) a personalized Federated Averaging scheme that weights client updates by local data volume. Evaluated on FER2013 and a custom physiological dataset, the federated Convolutional Neural Network attains 77% accuracy, the Random Forest 74%, and their fusion 87%, matching a centralized baseline while keeping all raw data local. The developed system converges in 18 rounds, with an average round time of 120 seconds and a per-client memory footprint below 200 MB. These results indicate that FedMultiEmo offers a practical approach to real-time, privacy-aware emotion recognition in automotive settings.

Figures

Figures reproduced from arXiv: 2507.15470 by the authors.

Figure 1
Figure 1. Pipeline for FedMultiEmo: A Multimodal Federated Learning Approach for Real-time Emotion Recognition, Illustrating [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Comparison of Confusion Matrices for Physiological, Image, and Multimodal Data in 7-Emotion Classification. [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 31 canonical work pages

  1. [1]

    Human-machine interaction and driver monitoring systems,

    F. I. for Integrated Circuits IIS, “Human-machine interaction and driver monitoring systems,” Fraunhofer IIS, 2023. Accessed: 2025-03-03

    work page 2023

  2. [2]

    Federated learning for comfort features in vehicles with collaborative sensing: A review,

    B. C. G ¨ul, D. Dittler, N. Jazdi, and M. Weyrich, “Federated learning for comfort features in vehicles with collaborative sensing: A review,” in 2024 IEEE 29th International Conference on Emerging Technologies and Factory Automation (ETFA), pp. 1–7, 2024

    work page 2024

  3. [3]

    Personalized comfort features in software-defined vehicles using federated learning,

    B. C. G ¨ul, N. Devarakonda, N. Jazdi, and M. Weyrich, “Personalized comfort features in software-defined vehicles using federated learning,” in 2024 IEEE 29th International Conference on Emerging Technologies and Factory Automation (ETFA), pp. 1–7, 2024

    work page 2024

  4. [4]

    Challenges in facial emotion recognition under real-world variability,

    Z. Xu, J. Li, and Y . Wang, “Challenges in facial emotion recognition under real-world variability,”IEEE Transactions on Affective Computing, vol. 12, no. 1, pp. 52–64, 2021

    work page 2021

  5. [5]

    Privacy and security in emotion recognition systems: A review,

    R. Sharma, S. Singh, and S. Kaur, “Privacy and security in emotion recognition systems: A review,” Journal of Privacy and Confidentiality , vol. 8, no. 2, pp. 135–150, 2020

    work page 2020

  6. [6]

    Federated learning for privacy-preserving emo- tion recognition from multimodal data,

    Q. Li and X. Yang, “Federated learning for privacy-preserving emo- tion recognition from multimodal data,” IEEE Transactions on Neural Networks and Learning Systems , vol. 32, no. 3, pp. 815–825, 2021

    work page 2021

  7. [7]

    Decentralized online federated g-network learning for lightweight intrusion detection,

    M. Nakip, B. C. G ¨ul, and E. Gelenbe, “Decentralized online federated g-network learning for lightweight intrusion detection,” in 2023 31st international symposium on modeling, analysis, and simulation of com- puter and telecommunication systems (MASCOTS), pp. 1–8, IEEE, 2023

    work page 2023

  8. [8]

    Using federated learning in the context of software-defined mobility systems for predictive quality of service,

    B. G ¨ul, N. Devarakonda, D. Dittler, N. Jazdi, and M. Weyrich, “Using federated learning in the context of software-defined mobility systems for predictive quality of service,” vol. 2419, 2023

    work page 2023

Show all 34 references
  1. [9]

    Affectnet: A database for facial expression, valence, and arousal computing,

    A. Mollahosseini, B. Hasani, and M. H. Mahoor, “Affectnet: A database for facial expression, valence, and arousal computing,” IEEE Transac- tions on Affective Computing , vol. 10, no. 1, pp. 18–31, 2019

    2019

  2. [10]

    Multimodal emotion recognition using deep learning: A survey,

    S. Ghosh, P. Sharma, and M. Gupta, “Multimodal emotion recognition using deep learning: A survey,” IEEE Access , vol. 7, pp. 173643– 173654, 2019

    2019

  3. [11]

    Fed-phyers: A federated learn- ing approach for multimodal emotion recognition using physiological signals,

    M. Sharma, R. Gupta, and S. Agarwal, “Fed-phyers: A federated learn- ing approach for multimodal emotion recognition using physiological signals,” IEEE Access, vol. 12, pp. 12564–12574, 2024

    2024

  4. [12]

    Automotive ai

    Affectiva, “Automotive ai.” https://www.affectiva.com, 2023. Accessed: 2023-03-03

    2023

  5. [13]

    Seeing machines driver monitoring technology: A real-world application,

    A. Tawari, H. Martin, and S. Larsen, “Seeing machines driver monitoring technology: A real-world application,” SAE International Journal of Connected and Automated Vehicles , vol. 4, no. 1, pp. 23–38, 2021

    2021

  6. [14]

    Active health monitor- ing solutions,

    Fraunhofer Institute for Integrated Circuits IIS, “Active health monitor- ing solutions,” 2023. Accessed: 2023-03-03

    2023

  7. [15]

    Real-world challenges in facial emotion recognition: A comprehensive survey,

    M. Zhao, J. Zhang, D. Li, and J. Sun, “Real-world challenges in facial emotion recognition: A comprehensive survey,” IEEE Transactions on Affective Computing, vol. 11, no. 2, pp. 380–394, 2020

    2020

  8. [16]

    Physiological signal analysis for emotion recognition using random forest and feature selection,

    S. Tripathi, A. Agrawal, and P. Verma, “Physiological signal analysis for emotion recognition using random forest and feature selection,” In- ternational Journal of Biomedical Engineering and Technology, vol. 37, no. 3, pp. 197–214, 2021. Preprint version. Accepted for publica...

    2021

  9. [17]

    Privacy and ethical concerns in in-vehicle emotion tracking technologies,

    J. Smith, K. Li, and Y . Zhang, “Privacy and ethical concerns in in-vehicle emotion tracking technologies,” Proceedings of the AAAI Conference on Artificial Intelligence, vol. 36, no. 1, pp. 4552–4561, 2022

    2022

  10. [18]

    Multimodal fusion of physio- logical and visual cues for real-time driver emotion detection,

    R. Verma, N. Kumar, and A. Prakash, “Multimodal fusion of physio- logical and visual cues for real-time driver emotion detection,” Neural Computing and Applications , vol. 34, no. 5, pp. 12789–12805, 2022

    2022

  11. [19]

    Emotion recognition using physiological signals: A review,

    J. Healey, S. M. Sharma, and R. Picard, “Emotion recognition using physiological signals: A review,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 41, no. 12, pp. 2961–2982, 2019

    2019

  12. [20]

    Facial expression recognition: A literature review,

    Z. Zeng, J. P. G. L. Hu, and S. J. Ho, “Facial expression recognition: A literature review,” IEEE Trans. Affective Computing , vol. 10, no. 3, pp. 313–327, 2019

    2019

  13. [21]

    End-to-end deep learning for emotion recognition in the wild,

    S. Ghosh et al. , “End-to-end deep learning for emotion recognition in the wild,” IEEE Transactions on Affective Computing , vol. 11, no. 4, pp. 558–570, 2020

    2020

  14. [22]

    Deep convolutional neural networks for emotion recognition: A survey,

    P. Gupta et al. , “Deep convolutional neural networks for emotion recognition: A survey,” IEEE Transactions on Neural Networks and Learning Systems, vol. 32, no. 5, pp. 1312–1325, 2021

    2021

  15. [23]

    Physiosync: Temporal and cross-modal contrastive learning inspired by physiological synchronization for eeg-based emotion recognition,

    K. Cui, J. Li, Y . Liu, X. Zhang, Z. Hu, and M. Wang, “Physiosync: Temporal and cross-modal contrastive learning inspired by physiological synchronization for eeg-based emotion recognition,” arXiv preprint arXiv:2504.17163, 2025

    2025 arXiv

  16. [24]

    Multimodal emotion recognition using visual, vocal and physiological signals: a review,

    G. Udahemuka, K. Djouani, and A. M. Kurien, “Multimodal emotion recognition using visual, vocal and physiological signals: a review,” Applied Sciences, vol. 14, no. 17, p. 8071, 2024

    2024

  17. [25]

    Communication-efficient learning of deep networks from decentralized data,

    B. McMahan, E. Moore, D. Ramage, S. Hampson, and B. A. y Arcas, “Communication-efficient learning of deep networks from decentralized data,” in Artificial intelligence and statistics , pp. 1273–1282, PMLR, 2017

    2017

  18. [26]

    Fed-phyers: Federated learning- based multi-modal emotion recognition using physiological signals,

    R. Sharma, A. Gupta, and M. Singh, “Fed-phyers: Federated learning- based multi-modal emotion recognition using physiological signals,” Multimedia Tools and Applications , 2024

    2024

  19. [27]

    Federated learning for eeg-based emotion recognition,

    M. Zhou, X. Liu, and Y . Zhang, “Federated learning for eeg-based emotion recognition,” Electronics, vol. 11, no. 20, p. 3316, 2022

    2022

  20. [28]

    Fedcmd: Federated cross-modal distillation for driver emotion recognition,

    J. Li, X. Wang, and H. Xu, “Fedcmd: Federated cross-modal distillation for driver emotion recognition,” Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, vol. 8, no. 1, pp. 1–23, 2024

    2024

  21. [29]

    Flower: A friendly federated learning research framework,

    D. J. Beutel, T. Topal, A. Mathur, X. Qiu, J. Fernandez-Marques, Y . Gao, L. Sani, K. H. Li, T. Parcollet, P. P. B. de Gusm ˜ao, et al. , “Flower: A friendly federated learning research framework,” arXiv preprint arXiv:2007.14390, 2020

    2007 arXiv

  22. [30]

    Challenges in representation learning: A report on three machine learning contests,

    I. J. Goodfellow, D. Erhan, P. Luc Carrier, A. Courville, M. Mirza, B. Hamner, W. Cukierski, Y . Tang, D. Thaler, D.-H. Lee, Y . Zhou, C. Ramaiah, F. Feng, R. Li, X. Wang, D. Athanasakis, J. Shawe-Taylor, M. Milakov, J. Park, R. Ionescu, M. Popescu, C. Grozea, J. Bergstra, J. ...

    2015

  23. [31]

    Training deep networks for facial expression recognition with crowd-sourced label distribution,

    E. Barsoum, C. Zhang, C. C. Ferrer, and Z. Zhang, “Training deep networks for facial expression recognition with crowd-sourced label distribution,” in Proceedings of the 18th ACM international conference on multimodal interaction , pp. 279–283, 2016

    2016

  24. [32]

    A hybrid framework for latency compensation in remote testing of automotive electronic control units,

    S. Tziampazis, P. Hirmer, and M. Weyrich, “A hybrid framework for latency compensation in remote testing of automotive electronic control units,” Frontiers in the Internet of Things , vol. V olume 3 - 2024, 2024

    2024

  25. [33]

    Syncfed: Time- aware federated learning through explicit timestamping and synchro- nization,

    B. C. G ¨ul, S. Tziampazis, N. Jazdi, and M. Weyrich, “Syncfed: Time- aware federated learning through explicit timestamping and synchro- nization,” 2025

    2025

  26. [34]

    Distributed integration of electronic control units for automotive oems: Challenges, vision, and research directions,

    S. Tziampazis, O. Kopp, and M. Weyrich, “Distributed integration of electronic control units for automotive oems: Challenges, vision, and research directions,” in 2023 IEEE 20th International Conference on Software Architecture Companion (ICSA-C) , pp. 296–300, 2023. Preprint ...

    2023

Pith tools

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