CoChat
Sign Up
PreprintCited by 15Privacy-Preserving Technologies in DataIoT and Edge/Fog ComputingDomain Adaptation and Few-Shot Learning

TinyFedTL: Federated Transfer Learning on Tiny Devices

Open access
Kopparapu, Kavya, Lin, Eric·2021·arXiv
DOI 10.48550/arxiv.2110.01107arXiv 2110.01107
Open this paper in CoChat — free accountOpen PDF

Abstract

TinyML has rose to popularity in an era where data is everywhere. However, the data that is in most demand is subject to strict privacy and security guarantees. In addition, the deployment of TinyML hardware in the real world has significant memory and communication constraints that traditional ML fails to address. In light of these challenges, we present TinyFedTL, the first implementation of federated transfer learning on a resource-constrained microcontroller.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

DataCite · 10.48550/arxiv.2110.01107

Cites

9 of 9 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Tiny-FedTL uses Arduino to deploy on top of TensorFlow Lite Micro with an on-device training facility.

    “It uses Arduino to deploy the TinyFedTL on top of Tensorflow Lite Micro with an on-device training facility.”

    From A review on TinyML: State-of-the-art and prospects (2022) · cites this paper as Kopparapu and Lin, 2021

    The Introduction (p. 1) identifies an Arduino implementation, and §4.4.2 (p. 6) reports training on-device. But §3.3 (p. 3) says TensorFlow Lite Micro was incompatible with on-device training, so the authors implemented their own training update in C++; the claim that TinyFedTL was deployed on top of TensorFlow Lite Micro with a training facility is not supported. (LLM, conf=0.88)

    Experimental — accuracy not yet measured

  • The Tiny-FedTL approach deploys a machine learning model without needing excess memory utilization.

    “The implied Tiny-FedTL approach deploys a machine learning model without needing excess memory utilization.”

    From A review on TinyML: State-of-the-art and prospects (2022) · cites this paper as Kopparapu and Lin, 2021

  • Kopparapu and Lin (2021) implement federated learning for IoT-enabled edge devices.

    “In (Kopparapu and Lin, 2021), implementation of federated learning is done for IoT-enabled edge devices.”

    From A review on TinyML: State-of-the-art and prospects (2022) · cites this paper as Kopparapu and Lin, 2021

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

1 of the 10 references couldn’t be matched to a known paper.

Research works printed but not linked: 3

  1. [3][3] Brendan McMahan, Eider Moore, Daniel Ramage, Seth Hampson, and Blaise Aguera y Arcas. Communicationefficient learning of deep networks from decentralized data. In Artificial Intelligence and Statistics , pages 12731282. PMLR, 2017. -
  2. [5][5] Maria Florina Balcan, Avrim Blum, Shai Fine, and Yishay Mansour. Distributed learning, communication complexity and privacy. In Conference on Learning Theory , pages 26-1, 2012. -
  3. [9][9] Han Cai, Chuang Gan, Ligeng Zhu, and Song Han. Tinytl: Reduce memory, not parameters for efficient ondevice learning. Advances in Neural Information Processing Systems , 33, 2020. -

Cited by

Sorted by citations

Related

Authors
Kopparapu, K., & Lin, E.
Affiliations
Harvard University (US)
Venue
arXiv
Publisher
arXiv
Published
2021
ISSN
2331-8422 (ISSN-L)
DOI
10.48550/arxiv.2110.01107
arXiv
2110.01107
OpenAlex
W3203576251
Cited by
14
References
10
Citations by year
2025: 1 · 2024: 6 · 2023: 5 · 2022: 2
Type
posted content
Language
en
Open access
Licence
creativecommons.org/licenses/by/4.0/legalcode
Locations
  • arXiv (Cornell University) (submittedVersion)
  • arXiv (Cornell University) (cc-by)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Privacy-Preserving Technologies in Data
  • Physical Sciences › Computer Science › Computer Networks and Communications › IoT and Edge/Fog Computing
  • Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
Keywords
federated transfer learning, TinyML, microcontroller, resource-constrained devices, edge computing, privacy-preserving machine learning

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review