mlsanityData to 14 Oct 2026

Avoid Overfitting User Specific Information in Federated Keyword Spotting

arXiv 2206.08864 ↗First posted 17 Jun 2022cs.LG cs.MM cs.SDcited by 4

Not on the map. Why?

Abstract (arXiv)

Keyword spotting (KWS) aims to discriminate a specific wake-up word from other signals precisely and efficiently for different users. Recent works utilize various deep networks to train KWS models with all users' speech data centralized without considering data privacy. Federated KWS (FedKWS) could serve as a solution without directly sharing users' data. However, the small amount of data, different user habits, and various accents could lead to fatal problems, e.g., overfitting or weight divergence. Hence, we propose several strategies to encourage the model not to overfit user-specific information in FedKWS. Specifically, we first propose an adversarial learning strategy, which updates the downloaded global model against an overfitted local model and explicitly encourages the global model to capture user-invariant information. Furthermore, we propose an adaptive local training strategy, letting clients with more training data and more uniform class distributions undertake more local update steps. Equivalently, this strategy could weaken the negative impacts of those users whose data is less qualified. Our proposed FedKWS-UI could explicitly and implicitly learn user-invariant information in FedKWS. Abundant experimental results on federated Google Speech Commands verify the effectiveness of FedKWS-UI.

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Builds on 30 of its 37 references that are in the map. First posted on arXiv 17 Jun 2022.

Builds on

30 of 37 references.

References · 30

  1. 2015Distilling the Knowledge in a Neural Networkcited by 10,324
  2. 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,669
  3. 2018Federated Optimization in Heterogeneous Networkscited by 2,987
  4. 2016Deep Learning with Differential Privacycited by 2,728
  5. 2014How transferable are features in deep neural networks?cited by 1,225
  6. 2018Federated Learning with Non-IID Datacited by 1,108
  7. 2020Adaptive Federated Optimizationcited by 1,046
  8. 2021Model-Contrastive Federated Learningcited by 766
  9. 2018Speech Commands: A Dataset for Limited-Vocabulary Speech Recognitioncited by 752
  10. 2021AST: Audio Spectrogram Transformercited by 643
  11. 2021Federated Learning Based on Dynamic Regularizationcited by 562
  12. 2015Listen, Attend and Spellcited by 543
18 more
  1. 2021Exploiting Shared Representations for Personalized Federated Learningcited by 541
  2. 2019Federated Machine Learning: Concept and Applicationscited by 412
  3. 2020Think Locally, Act Globally: Federated Learning with Local and Global Representationscited by 325
  4. 2021FedScale: Benchmarking Model and System Performance of Federated Learning at Scalecited by 136
  5. 2017Hello Edge: Keyword Spotting on Microcontrollerscited by 110
  6. 2018Federated Learning for Keyword Spottingcited by 95
  7. 2020Federated Mutual Learningcited by 71
  8. 2017Deep Residual Learning for Small-Footprint Keyword Spottingcited by 68
  9. 2022Personalized Federated Learning via Variational Bayesian Inferencecited by 68
  10. 2021Keyword Transformer: A Self-Attention Model for Keyword Spottingcited by 60
  11. 2020Streaming keyword spotting on mobile devicescited by 47
  12. 2020Training Speech Recognition Models with Federated Learning: A Quality/Cost Frameworkcited by 46
  13. 2020Training Keyword Spotting Models on Non-IID Data with Federated Learningcited by 36
  14. 2022Federated Learning with Position-Aware Neuronscited by 24
  15. 2021Federated Acoustic Modeling For Automatic Speech Recognitioncited by 20
  16. 2021SubSpectral Normalization for Neural Audio Data Processingcited by 11
  17. 2016The Microsoft 2016 Conversational Speech Recognition Systemcited by 9
  18. 2021Aggregate or Not? Exploring Where to Privatize in DNN Based Federated Learning Under Different Non-IID Scenescited by 4

Built on by

Cited by year

Citing papers per year: 2023: 2; 2024: 2.

4 papers.

  1. 2024MAP: Model Aggregation and Personalization in Federated Learning with Incomplete Classescited by 7
  2. 2023MrTF: Model Refinery for Transductive Federated Learningcited by 2
  3. 2024Towards Federated Learning with On-device Training and Communication in 8-bit Floating Pointcited by 1
  4. 2023DCFL: Non-IID awareness Data Condensation aided Federated Learningcited by 0