Pith. sign in

REVIEW 2 major objections 1 minor 42 references

Devices share inference feedback to learn optimal DNN partitions, cutting multi-user edge latency by up to 50%.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · grok-4.3

2026-06-27 17:20 UTC pith:2HW5WZFJ

load-bearing objection CANS adds device grouping plus offline warm-start to a federated LinUCB variant for multi-user edge DNN partitioning and supplies a regret bound, yet the 50% latency claim rests on a two-device prototype whose transfer assumptions are untested under real channel variation. the 2 major comments →

arxiv 2606.09175 v1 pith:2HW5WZFJ submitted 2026-06-08 cs.LG cs.AIcs.DC

CANS: Accelerating Multiuser Collaborative Edge Inference via Cooperative Autodidactic NeuroSurgeon

Zheshun Wu , Ziyang Zhang , Changyao Lin , Zenglin Xu , Jie Liu This is my paper
classification cs.LG cs.AIcs.DC
keywords collaborative edge inferenceDNN partitioningonline learningfederated banditmulti-user MECinference latencyedge computing
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper proposes CANS as a framework in which multiple mobile devices run collaborative DNN inference by partitioning each model and offloading backend parts to a shared edge server. Devices exchange feedback on partition choices during live operation so that each can improve its split decision without central coordination. A new FedLinUCB-DW algorithm groups devices of matching type and initializes online search from each device's local offline early-exit records. The resulting system is shown to produce lower average inference latency than non-cooperative baselines on both simulation and two-device hardware tests, while also carrying a derived regret bound. A sympathetic reader would care because faster on-device intelligence becomes feasible when devices can pool their learning signals under changing wireless conditions.

Core claim

CANS enables devices to adaptively learn optimal DNN partitions by sharing informative feedback during online inference. To handle device heterogeneity and leverage offline experience, the framework integrates FedLinUCB-DW, which groups devices of the same type and warm-starts online exploration from local offline early-exit inference experience. Theoretical analysis supplies a regret upper bound for FedLinUCB-DW. Prototype experiments on two edge devices show that CANS reduces average inference latency by up to 50% relative to the non-cooperative baseline.

What carries the argument

The CANS framework together with the FedLinUCB-DW algorithm that groups same-type devices and warm-starts online partition search from offline early-exit data.

Load-bearing premise

Devices of the same type can be reliably grouped and their local offline early-exit records transfer usefully to warm-start online learning when wireless conditions and device capabilities vary over time.

What would settle it

A controlled test on multiple same-type devices in which the FedLinUCB-DW warm-start version produces higher cumulative latency or larger regret than a version started from random initialization under the same time-varying wireless traces.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Multi-user collaborative inference achieves lower average latency than independent per-device decisions.
  • Grouping by device type improves sample efficiency of the online partition search.
  • Offline early-exit traces accelerate convergence of the online learning process.
  • The learning procedure admits a finite regret upper bound that scales with the number of partition choices.

Where Pith is reading between the lines

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

  • The same grouping-plus-warm-start pattern could be tested on tasks other than DNN partitioning, such as joint resource allocation across devices.
  • If device-type labels become unreliable, performance may fall back toward the non-cooperative baseline.
  • Scaling the number of devices may require adjustments to the grouping step to keep communication overhead low.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

2 major / 1 minor

Summary. The paper proposes CANS, a framework for multi-user collaborative edge DNN inference over wireless MEC that enables devices to cooperatively learn optimal model partitions online via shared feedback. It integrates FedLinUCB-DW to group same-type devices and warm-start from offline early-exit runs, derives a regret upper bound for the algorithm, and reports latency reductions versus baselines in both simulation and a two-device hardware prototype (up to 50% vs. non-cooperative baseline).

Significance. If the offline-to-online transfer and regret bound hold under realistic non-stationary wireless conditions, the work would provide a practical mechanism for adaptive collaborative inference that reduces latency in heterogeneous edge settings while offering theoretical guarantees.

major comments (2)
  1. [Abstract and description of FedLinUCB-DW integration] The central empirical claim of up to 50% latency reduction in the two-device prototype (Abstract) rests on the untested assumption that local offline early-exit experience transfers effectively to warm-start FedLinUCB-DW under time-varying wireless channels and device heterogeneity; the prototype provides no direct measurement of this transfer or of performance with larger device counts.
  2. [Theoretical guarantees section for FedLinUCB-DW] The regret upper bound for FedLinUCB-DW is derived under standard linear contextual bandit assumptions (stationary contexts), yet the system model includes time-varying wireless links that induce non-stationary context distributions; no analysis or extension addresses whether the bound remains valid or how the algorithm adapts.
minor comments (1)
  1. [Abstract] The abstract states results on 'a simulated environment and a hardware prototype system' but provides no details on experimental controls, number of runs, or statistical significance for the latency figures.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments on our manuscript. We address each major comment below, indicating planned revisions where appropriate.

read point-by-point responses

  1. Referee: The central empirical claim of up to 50% latency reduction in the two-device prototype (Abstract) rests on the untested assumption that local offline early-exit experience transfers effectively to warm-start FedLinUCB-DW under time-varying wireless channels and device heterogeneity; the prototype provides no direct measurement of this transfer or of performance with larger device counts.

    Authors: We acknowledge that the two-device hardware prototype does not isolate the offline-to-online transfer effect via ablation nor include results for device counts beyond two. The reported latency reduction reflects end-to-end CANS performance (including warm-start) in the prototype environment. Simulation results with heterogeneous device groups provide supporting evidence for the approach under varying conditions. We will add an explicit limitations paragraph in the experimental section noting the prototype scale and the implicit evaluation of transfer, along with suggestions for future larger-scale hardware validation. This constitutes a partial revision. revision: partial

  2. Referee: The regret upper bound for FedLinUCB-DW is derived under standard linear contextual bandit assumptions (stationary contexts), yet the system model includes time-varying wireless links that induce non-stationary context distributions; no analysis or extension addresses whether the bound remains valid or how the algorithm adapts.

    Authors: The regret upper bound is derived under the standard stationary-context assumptions of linear contextual bandits, as stated in the theoretical analysis section. The FedLinUCB-DW algorithm adapts to observed feedback in the online phase, which empirically handles time-varying wireless conditions in both simulations and the prototype. We do not provide a non-stationary regret analysis, which would require substantial additional theoretical development. We will revise the theoretical guarantees section to explicitly restate the stationarity assumption and discuss its implications for applicability under strong non-stationarity. This is a partial revision. revision: partial

Circularity Check

0 steps flagged

No circularity detected in regret bound derivation or empirical claims

full rationale

The paper derives a regret upper bound for FedLinUCB-DW as an independent theoretical guarantee and reports empirical latency reductions from separate prototype and simulation experiments. No claimed result reduces by construction to a fitted parameter, self-citation, or input definition; the offline-to-online transfer is an explicit modeling assumption rather than a self-referential step. The derivation chain remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review supplies no explicit free parameters, axioms, or invented entities; the regret bound and algorithm are described at high level without listing fitted constants or unproven background assumptions.

pith-pipeline@v0.9.1-grok · 5795 in / 1154 out tokens · 18941 ms · 2026-06-27T17:20:59.542852+00:00 · methodology

0 comments
read the original abstract

Recently, mobile edge computing (MEC)-enabled collaborative deep neural network (DNN) inference has emerged as a promising approach for delivering intelligent services to resource-constrained mobile devices. A representative scenario is multi-user collaborative edge inference, where distinct devices independently partition their DNN models and offload backend computation to a common edge server over wireless networks. However, determining the optimal DNN partition for each device is challenging due to unknown and time-varying system conditions, including fluctuating wireless links and diverse device capabilities. To address this problem, we propose Cooperative Autodidactic NeuroSurgeon (CANS), a collaborative edge inference framework that enables devices to adaptively learn optimal DNN partitions by sharing informative feedback during online inference. To handle the challenge of device heterogeneity and better leverage offline inference experience, we integrate a novel FedLinUCB-DW algorithm that groups devices of the same type and warm-starts online exploration using local offline early-exit inference experience. Furthermore, we provide theoretical guarantees for FedLinUCB-DW by deriving the regret upper bound. We also validate our method on both a simulated environment and a hardware prototype system. Empirical evaluations demonstrate that CANS achieves lower inference latency compared to state-of-the-art baselines. Especially, in prototype experiments on two edge devices, the proposed CANS reduced average inference latency by up to 50% compared to the non-cooperative baseline.

Figures

Figures reproduced from arXiv: 2606.09175 by Changyao Lin, Jie Liu, Zenglin Xu, Zheshun Wu, Ziyang Zhang.

Figure 1
Figure 1. Figure 1: The system architecture of CANS. scenario by jointly optimizing DNN partitioning and server￾resource allocation [19], [20]. However, they still consider conducting offline profiling under the restrictive assumption that system parameters are known in advance, which is impractical in realistic deployments. The work most closely related to ours is [21], which iteratively solves optimization problems to ident… view at source ↗
Figure 2
Figure 2. Figure 2: FedLinUCB for asynchronous multiuser collaborative edge inference. [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: The workflow of FedLinUCB-DW. B. FedLinUCB with Device Grouping and Local Warm-Start Given the limitations of the original FedLinUCB outlined above, we devise a novel distributed online learning algo￾rithm, called FedLinUCB with Device Grouping and Local Warm-Start (FedLinUCB-DW) to support our proposed CANS framework, which addresses these shortcomings based on theoretically grounded mechanisms. Heterogen… view at source ↗
Figure 4
Figure 4. Figure 4: Illustration of device grouping. The server uses device [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Cumulative Regret for three backbone models. [PITH_FULL_IMAGE:figures/full_fig_p011_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Average latency for three backbone models. [PITH_FULL_IMAGE:figures/full_fig_p011_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Estimation error for three backbone models. [PITH_FULL_IMAGE:figures/full_fig_p011_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Average latency of different numbers of agents for three backbone models. [PITH_FULL_IMAGE:figures/full_fig_p012_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Impact of computational heterogeneity within each [PITH_FULL_IMAGE:figures/full_fig_p012_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Overall hardware prototype testbed of CANS. [PITH_FULL_IMAGE:figures/full_fig_p013_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Average latency for different baselines on Xavier NX. [PITH_FULL_IMAGE:figures/full_fig_p014_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Average latency for different baselines on Jetson Orin Nano. [PITH_FULL_IMAGE:figures/full_fig_p014_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Average latency of ablation study on Xavier NX. [PITH_FULL_IMAGE:figures/full_fig_p014_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Average latency of ablation study on Jetson Orin Nano. [PITH_FULL_IMAGE:figures/full_fig_p014_14.png] view at source ↗

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

42 extracted references · 1 canonical work pages

  1. [1]

    Cloud versus edge deployment strategies of real-time face recognition inference,

    A. Koubaa, A. Ammar, A. Kanhouch, and Y . AlHabashi, “Cloud versus edge deployment strategies of real-time face recognition inference,” IEEE Transactions on Network Science and Engineering, vol. 9, no. 1, pp. 143–160, 2021

    2021

  2. [2]

    Edge video analytics: A survey on applications, systems and enabling techniques,

    R. Xu, S. Razavi, and R. Zheng, “Edge video analytics: A survey on applications, systems and enabling techniques,”IEEE Communications Surveys & Tutorials, vol. 25, no. 4, pp. 2951–2982, 2023

    2023

  3. [3]

    Survey on virtual assistant: Google assistant, siri, cortana, alexa,

    A. S. Tulshan and S. N. Dhage, “Survey on virtual assistant: Google assistant, siri, cortana, alexa,” inInternational symposium on signal processing and intelligent recognition systems, pp. 190–201, Springer, 2018

    2018

  4. [4]

    Dynamic computation offloading for mobile-edge computing with energy harvesting devices,

    Y . Mao, J. Zhang, and K. B. Letaief, “Dynamic computation offloading for mobile-edge computing with energy harvesting devices,”IEEE Journal on Selected Areas in Communications, vol. 34, no. 12, pp. 3590– 3605, 2016

    2016

  5. [5]

    Efficient acceleration of deep learning inference on resource-constrained edge devices: A review,

    M. M. H. Shuvo, S. K. Islam, J. Cheng, and B. I. Morshed, “Efficient acceleration of deep learning inference on resource-constrained edge devices: A review,”Proceedings of the IEEE, vol. 111, no. 1, pp. 42– 91, 2022

    2022

  6. [6]

    Computation offloading in resource-constrained multi-access edge computing,

    K. Li, X. Wang, Q. He, J. Wang, J. Li, S. Zhan, G. Lu, and S. Dust- dar, “Computation offloading in resource-constrained multi-access edge computing,”IEEE Transactions on Mobile Computing, vol. 23, no. 11, pp. 10665–10677, 2024

    2024

  7. [7]

    Learning- aided computation offloading for trusted collaborative mobile edge computing,

    Y . Li, X. Wang, X. Gan, H. Jin, L. Fu, and X. Wang, “Learning- aided computation offloading for trusted collaborative mobile edge computing,”IEEE Transactions on Mobile Computing, vol. 19, no. 12, pp. 2833–2849, 2020

    2020

  8. [8]

    Delay- aware microservice coordination in mobile edge computing: A rein- forcement learning approach,

    S. Wang, Y . Guo, N. Zhang, P. Yang, A. Zhou, and X. Shen, “Delay- aware microservice coordination in mobile edge computing: A rein- forcement learning approach,”IEEE Transactions on Mobile Computing, vol. 20, no. 3, pp. 939–951, 2021

    2021

  9. [9]

    Deep reinforcement learning for online computation offloading in wireless powered mobile-edge computing networks,

    L. Huang, S. Bi, and Y .-J. A. Zhang, “Deep reinforcement learning for online computation offloading in wireless powered mobile-edge computing networks,”IEEE Transactions on Mobile Computing, vol. 19, no. 11, pp. 2581–2593, 2020

    2020

  10. [10]

    Computation offloading scheduling for periodic tasks in mobile edge computing,

    S. Jo ˇsilo and G. D ´an, “Computation offloading scheduling for periodic tasks in mobile edge computing,”IEEE/ACM Transactions on Network- ing, vol. 28, no. 2, pp. 667–680, 2020

    2020

  11. [11]

    Neurosurgeon: Collaborative intelligence between the cloud and mobile edge,

    Y . Kang, J. Hauswald, C. Gao, A. Rovinski, T. Mudge, J. Mars, and L. Tang, “Neurosurgeon: Collaborative intelligence between the cloud and mobile edge,”ACM SIGARCH Computer Architecture News, vol. 45, no. 1, pp. 615–629, 2017

    2017

  12. [12]

    A survey on collaborative dnn inference for edge intelligence,

    W.-Q. Ren, Y .-B. Qu, C. Dong, Y .-Q. Jing, H. Sun, Q.-H. Wu, and S. Guo, “A survey on collaborative dnn inference for edge intelligence,” Machine Intelligence Research, vol. 20, no. 3, pp. 370–395, 2023

    2023

  13. [13]

    Autodidactic neurosurgeon: Collaborative deep inference for mobile edge intelligence via online learning,

    L. Zhang, L. Chen, and J. Xu, “Autodidactic neurosurgeon: Collaborative deep inference for mobile edge intelligence via online learning,” in Proceedings of the Web Conference 2021, pp. 3111–3123, 2021

    2021

  14. [14]

    Dynamic adaptive dnn surgery for inference acceleration on the edge,

    C. Hu, W. Bao, D. Wang, and F. Liu, “Dynamic adaptive dnn surgery for inference acceleration on the edge,” inIEEE INFOCOM 2019 - IEEE Conference on Computer Communications, pp. 1423–1431, 2019

    2019

  15. [15]

    Distributed assignment with load balancing for dnn inference at the edge,

    Y . Xu, T. Mohammed, M. Di Francesco, and C. Fischione, “Distributed assignment with load balancing for dnn inference at the edge,”IEEE Internet of Things Journal, vol. 10, no. 2, pp. 1053–1065, 2023

    2023

  16. [16]

    Learning the optimal partition for collaborative dnn training with privacy requirements,

    L. Zhang and J. Xu, “Learning the optimal partition for collaborative dnn training with privacy requirements,”IEEE Internet of Things Journal, vol. 9, no. 13, pp. 11168–11178, 2022

    2022

  17. [17]

    Adversarial group linear bandits and its application to collaborative edge inference,

    Y . Huang, L. Zhang, and J. Xu, “Adversarial group linear bandits and its application to collaborative edge inference,” inIEEE INFOCOM 2023- IEEE Conference on Computer Communications, pp. 1–10, IEEE, 2023

    2023

  18. [18]

    Learning the optimal path and dnn partition for collaborative edge inference,

    Y . Huang, L. Zhang, and J. Xu, “Learning the optimal path and dnn partition for collaborative edge inference,”IEEE Transactions on Mobile Computing, pp. 1–14, 2025

    2025

  19. [19]

    Task partitioning and offloading in dnn-task enabled mobile edge computing networks,

    M. Gao, R. Shen, L. Shi, W. Qi, J. Li, and Y . Li, “Task partitioning and offloading in dnn-task enabled mobile edge computing networks,” IEEE Transactions on Mobile Computing, vol. 22, no. 4, pp. 2435–2445, 2023

    2023

  20. [20]

    Joint multiuser dnn partitioning and computational resource allocation for collaborative edge intelligence,

    X. Tang, X. Chen, L. Zeng, S. Yu, and L. Chen, “Joint multiuser dnn partitioning and computational resource allocation for collaborative edge intelligence,”IEEE Internet of Things Journal, vol. 8, no. 12, pp. 9511– 9522, 2021

    2021

  21. [21]

    Multi-agent systems for collaborative inference based on deep policy q-inference network,

    S. Wang, Y . Jing, K. Wang, and X. Wang, “Multi-agent systems for collaborative inference based on deep policy q-inference network,” Journal of Grid Computing, vol. 22, no. 1, p. 38, 2024. 16

    2024

  22. [22]

    Distributed dnn inference with fine-grained model partitioning in mobile edge computing networks,

    H. Li, X. Li, Q. Fan, Q. He, X. Wang, and V . C. M. Leung, “Distributed dnn inference with fine-grained model partitioning in mobile edge computing networks,”IEEE Transactions on Mobile Computing, vol. 23, no. 10, pp. 9060–9074, 2024

    2024

  23. [23]

    Dronebandit: Multi-armed contextual bandits for collaborative edge-to-cloud inference in resource-constrained nanodrones,

    G. Chacun, M. Akeddar, T. Rieder, B. Da Rocha Carvalho, and M. Za- pater, “Dronebandit: Multi-armed contextual bandits for collaborative edge-to-cloud inference in resource-constrained nanodrones,” inPro- ceedings of the Great Lakes Symposium on VLSI 2024, pp. 98–104, 2024

    2024

  24. [24]

    Lattimore and C

    T. Lattimore and C. Szepesv ´ari,Bandit algorithms. Cambridge Univer- sity Press, 2020

    2020

  25. [25]

    Regret analysis of stochastic and nonstochastic multi-armed bandit problems,

    S. Bubeck, N. Cesa-Bianchi,et al., “Regret analysis of stochastic and nonstochastic multi-armed bandit problems,”Foundations and Trends® in Machine Learning, vol. 5, no. 1, pp. 1–122, 2012

    2012

  26. [26]

    Stochastic linear optimization under bandit feedback,

    V . Dani, T. P. Hayes, and S. M. Kakade, “Stochastic linear optimization under bandit feedback,” in21st Annual Conference on Learning Theory, no. 101, pp. 355–366, 2008

    2008

  27. [27]

    Improved algorithms for linear stochastic bandits,

    Y . Abbasi-Yadkori, D. P´al, and C. Szepesv ´ari, “Improved algorithms for linear stochastic bandits,”Advances in neural information processing systems, vol. 24, 2011

    2011

  28. [28]

    Distributed bandit learning: Near-optimal regret with efficient communication,

    Y . Wang, J. Hu, X. Chen, and L. Wang, “Distributed bandit learning: Near-optimal regret with efficient communication,” inICLR, OpenRe- view.net, 2020

    2020

  29. [29]

    A simple and provably efficient al- gorithm for asynchronous federated contextual linear bandits,

    J. He, T. Wang, Y . Min, and Q. Gu, “A simple and provably efficient al- gorithm for asynchronous federated contextual linear bandits,”Advances in neural information processing systems, vol. 35, pp. 4762–4775, 2022

    2022

  30. [30]

    Federated linear contextual bandits with heterogeneous clients,

    E. Blaser, C. Li, and H. Wang, “Federated linear contextual bandits with heterogeneous clients,” inInternational Conference on Artificial Intelligence and Statistics, pp. 631–639, PMLR, 2024

    2024

  31. [31]

    Thompson sampling for contextual bandits with linear payoffs,

    S. Agrawal and N. Goyal, “Thompson sampling for contextual bandits with linear payoffs,” inICML (3), JMLR Workshop and Conference Proceedings, pp. 127–135, JMLR.org, 2013

    2013

  32. [32]

    Parallel detection for efficient video analytics at the edge,

    Y . Wu, L. Liu, and R. Kompella, “Parallel detection for efficient video analytics at the edge,” inCogMI, pp. 1–10, IEEE, 2021

    2021

  33. [33]

    Real-time object detection for unmanned aerial vehicles based on vision transformer and edge computing,

    W. Zhu and K. Chen, “Real-time object detection for unmanned aerial vehicles based on vision transformer and edge computing,”Scientific Reports, 2026

    2026

  34. [34]

    Asynchronous upper confidence bound algorithms for federated linear bandits,

    C. Li and H. Wang, “Asynchronous upper confidence bound algorithms for federated linear bandits,” inInternational Conference on Artificial Intelligence and Statistics, pp. 6529–6553, PMLR, 2022

    2022

  35. [35]

    Joint optimization with dnn partitioning and resource allocation in mobile edge computing,

    C. Dong, S. Hu, X. Chen, and W. Wen, “Joint optimization with dnn partitioning and resource allocation in mobile edge computing,”IEEE Transactions on Network and Service Management, vol. 18, no. 4, pp. 3973–3986, 2021

    2021

  36. [36]

    Federated online clustering of bandits,

    X. Liu, H. Zhao, T. Yu, S. Li, and J. C. Lui, “Federated online clustering of bandits,” inUncertainty in Artificial Intelligence, pp. 1221–1231, PMLR, 2022

    2022

  37. [37]

    Early-exit deep neural network - A comprehensive survey,

    H. R. P, V . Srivastava, K. Chaurasia, R. G. Pacheco, and R. S. Couto, “Early-exit deep neural network - A comprehensive survey,”ACM Comput. Surv., vol. 57, no. 3, pp. 75:1–75:37, 2025

    2025

  38. [38]

    Adaee: Adaptive early-exit DNN inference through multi-armed bandits,

    R. G. Pacheco, M. Shifrin, R. D. S. Couto, D. S. Menasch ´e, M. K. Hanawal, and M. E. M. Campista, “Adaee: Adaptive early-exit DNN inference through multi-armed bandits,” inICC, pp. 3726–3731, IEEE, 2023

    2023

  39. [39]

    Cutting to the chase with warm-start contextual bandits,

    B. Oetomo, R. M. Perera, R. Borovica-Gajic, and B. I. P. Rubinstein, “Cutting to the chase with warm-start contextual bandits,”Knowl. Inf. Syst., vol. 65, no. 9, pp. 3533–3565, 2023

    2023

  40. [40]

    Run your visual- inertial odometry on nvidia jetson: Benchmark tests on a micro aerial vehicle,

    J. Jeon, S. Jung, E. Lee, D. Choi, and H. Myung, “Run your visual- inertial odometry on nvidia jetson: Benchmark tests on a micro aerial vehicle,”IEEE robotics and automation letters, vol. 6, no. 3, pp. 5332– 5339, 2021

    2021

  41. [41]

    Deepedgebench: Benchmarking deep neural networks on edge devices,

    S. P. Baller, A. Jindal, M. Chadha, and M. Gerndt, “Deepedgebench: Benchmarking deep neural networks on edge devices,”CoRR, vol. abs/2108.09457, 2021

    work page arXiv 2021

  42. [42]

    Netscope analyzer

    D. Schwendener, “Netscope analyzer.” http://dgschwend.github.io/ netscope/, 2016. 17 Supplementary Material In this supplementary material, we provide detailed proofs of the regret upper bound and communication complexity of the proposed FedLinUCB-DW algorithm. The overall analysis can be decomposed into two parts, corresponding to the front-end and back-...

    2016