mlsanityData to 7 Oct 2026

Training Speech Recognition Models with Federated Learning: A Quality/Cost Framework

arXiv 2010.15965 ↗First posted 29 Oct 2020cs.LG cs.DCcited by 46

Not on the map. Why?

Abstract (arXiv)

We propose using federated learning, a decentralized on-device learning paradigm, to train speech recognition models. By performing epochs of training on a per-user basis, federated learning must incur the cost of dealing with non-IID data distributions, which are expected to negatively affect the quality of the trained model. We propose a framework by which the degree of non-IID-ness can be varied, consequently illustrating a trade-off between model quality and the computational cost of federated training, which we capture through a novel metric. Finally, we demonstrate that hyper-parameter optimization and appropriate use of variational noise are sufficient to compensate for the quality impact of non-IID distributions, while decreasing the cost.

Key facts. Cited by 46 papers in the map (data through 7 Oct 2026). First posted on arXiv 29 Oct 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 14; 2022: 13; 2023: 8; 2024: 7; 2025: 2; 2026: 2.

46 papers.

Sort:

  1. 2024Automatic Speech Recognition using Advanced Deep Learning Approaches: A surveycited by 50
  2. 2022pFL-Bench: A Comprehensive Benchmark for Personalized Federated Learningcited by 44
  3. 2021Federated Learning Meets Natural Language Processing: A Surveycited by 42
  4. 2023Backdoor Attacks and Defenses in Federated Learning: Survey, Challenges and Future Research Directionscited by 33
  5. 2021End-to-End Speech Recognition from Federated Acoustic Modelscited by 30
  6. 2022A Review of Speech-centric Trustworthy Machine Learning: Privacy, Safety, and Fairnesscited by 13
  7. 2021Auto-weighted Robust Federated Learning with Corrupted Data Sourcescited by 12
  8. 2021Enabling On-Device Training of Speech Recognition Models with Federated Dropoutcited by 12
  9. 2024Training Machine Learning models at the Edge: A Surveycited by 11
  10. 2021Privacy attacks for automatic speech recognition acoustic models in a federated learning frameworkcited by 10
  11. 2021A Method to Reveal Speaker Identity in Distributed ASR Training, and How to Counter Itcited by 8
  12. 2021A Hybrid Architecture for Federated and Centralized Learningcited by 8
34 more
  1. 2021Separate but Together: Unsupervised Federated Learning for Speech Enhancement from Non-IID Datacited by 8
  2. 2021AdaSplit: Adaptive Trade-offs for Resource-constrained Distributed Deep Learningcited by 8
  3. 2022Federated Pruning: Improving Neural Network Efficiency with Federated Learningcited by 8
  4. 2023Importance of Smoothness Induced by Optimizers in FL4ASR: Towards Understanding Federated Learning for End-to-End ASRcited by 8
  5. 2024Communication-Efficient Personalized Federated Learning for Speech-to-Text Taskscited by 8
  6. 2022Federated Domain Adaptation for ASR with Full Self-Supervisioncited by 6
  7. 2022Decoupled Federated Learning for ASR with Non-IID Datacited by 6
  8. 2022Federated Self-Supervised Learning for Acoustic Event Classificationcited by 5
  9. 2022FedNST: Federated Noisy Student Training for Automatic Speech Recognitioncited by 5
  10. 2023Personalized Federated Learning on Long-Tailed Data via Adversarial Feature Augmentationcited by 5
  11. 2021Zero-Shot Federated Learning with New Classes for Audio Classificationcited by 4
  12. 2021Private Language Model Adaptation for Speech Recognitioncited by 4
  13. 2022Personalized Federated Learning with Exact Stochastic Gradient Descentcited by 4
  14. 2022Avoid Overfitting User Specific Information in Federated Keyword Spottingcited by 4
  15. 2022ILASR: Privacy-Preserving Incremental Learning for Automatic Speech Recognition at Production Scalecited by 4
  16. 2024Analyzing Inference Privacy Risks Through Gradients in Machine Learningcited by 4
  17. 2021Incremental Layer-wise Self-Supervised Learning for Efficient Speech Domain Adaptation On Devicecited by 3
  18. 2022Improving Automatic Speech Recognition for Non-Native English with Transfer Learning and Language Model Decodingcited by 3
  19. 2021Retrieving Speaker Information from Personalized Acoustic Models for Speech Recognitioncited by 2
  20. 2022Adaptive Multi-Corpora Language Model Training for Speech Recognitioncited by 2
  21. 2023RecUP-FL: Reconciling Utility and Privacy in Federated Learning via User-configurable Privacy Defensecited by 2
  22. 2021Exploring Heterogeneous Characteristics of Layers in ASR Models for More Efficient Trainingcited by 1
  23. 2022Group Personalized Federated Learningcited by 1
  24. 2023Speech Privacy Leakage from Shared Gradients in Distributed Learningcited by 1
  25. 2024HeteroSwitch: Characterizing and Taming System-Induced Data Heterogeneity in Federated Learningcited by 1
  26. 2024Personalized Federated Learning Techniques: Empirical Analysiscited by 1
  27. 2025Regularized Federated Learning for Privacy-Preserving Dysarthric and Elderly Speech Recognitioncited by 1
  28. 2023Ed-Fed: A generic federated learning framework with resource-aware client selection for edge devicescited by 0
  29. 2023Federated Representation Learning for Automatic Speech Recognitioncited by 0
  30. 2023Federated Active Learning for Target Domain Generalisationcited by 0
  31. 2024Federated Learning of Large ASR Models in the Real Worldcited by 0
  32. 2025Efficient and Privacy-Preserved Link Prediction via Condensed Graphscited by 0
  33. 2026Towards Personalized Federated Learning for Dysarthric Speech Recognitioncited by 0
  34. 2026A Practical Recipe for Semi-Supervised Federated ASR: Online Pseudo-Labels with Server Update Stabilizationcited by 0