The DOI matches its Crossref record.
Few-Shot Keyword Spotting in Any Language
Open accessAbstract
We introduce a few-shot transfer learning method for keyword spotting in any language. Leveraging open speech corpora in nine languages, we automate the extraction of a large multilingual keyword bank and use it to train an embedding model. With just five training examples, we fine-tune the embedding model for keyword spotting and achieve an average F1 score of 0.75 on keyword classification for 180 new keywords unseen by the embedding model in these nine languages. This embedding model also generalizes to new languages. We achieve an average F1 score of 0.65 on 5-shot models for 260 keywords sampled across 13 new languages unseen by the embedding model. We investigate streaming accuracy for our 5-shot models in two contexts: keyword spotting and keyword search. Across 440 keywords in 22 languages, we achieve an average streaming keyword spotting accuracy of 87.4% with a false acceptance rate of 4.3%, and observe promising initial results on keyword search.
Overview
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.
Integrity signals
Registry
Cites
Statistics
Verification
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.
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?
Mazumder et al. (2021) deploy a few-shot KWS that uses only 5 training examples and could be used for any language.
“A few-shot KWS is deployed in (Mazumder et al., 2021) that uses only 5 training examples which could be used for any language.”
From A review on TinyML: State-of-the-art and prospects (2022) · cites this paper as Mazumder et al., 2021
Mazumder et al. (2021) obtain an F1 score of 0.75 for 180 new keywords for 9 different languages.
“The procedure obtains an F1 score of 0.75 for 180 new keywords for 9 different languages.”
From A review on TinyML: State-of-the-art and prospects (2022) · cites this paper as Mazumder et al., 2021
Mazumder et al. (2021) find keyword spotting accuracy of 87.4% for 22 languages with 440 keywords.
“It is found that keyword spotting accuracy is 87.4% for 22 languages with 440 keywords.”
From A review on TinyML: State-of-the-art and prospects (2022) · cites this paper as Mazumder et al., 2021
The Mazumder et al. (2021) embedding model achieves an F1 score of 0.65 on just 5 shots.
“Such an embedding model can achieve a reasonable F 1 score of 0.65 on just 5 shots.”
From A review on TinyML: State-of-the-art and prospects (2022) · cites this paper as Mazumder et al., 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.
Citations
Cites
No references could be resolved yet.
Research works printed but not linked: 14
- [1][1] P. Warden, 'Speech commands: A dataset for limited-vocabulary speech recognition,' arXiv preprint arXiv:1804.03209 , 2018. - · arXiv
- [2][2] R. Ardila, M. Branson, K. Davis, M. Henretty, M. Kohler, J. Meyer, R. Morais, L. Saunders, F. M. Tyers, and G. Weber, 'Common voice: A massively-multilingual speech corpus,' arXiv preprint arXiv:1912.06670 , 2019. - · arXiv
- [4][4] G. Chen, C. Parada, and G. Heigold, 'Small-footprint keyword spotting using deep neural networks,' in 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 2014, pp. 4087-4091. -
- [5][5] T. N. Sainath and C. Parada, 'Convolutional neural networks for small-footprint keyword spotting,' in Sixteenth Annual Conference of the International Speech Communication Association , 2015. -
- [6][6] Y. Zhang, N. Suda, L. Lai, and V. Chandra, 'Hello edge: Keyword spotting on microcontrollers,' arXiv preprint arXiv:1711.07128 , 2017. - · arXiv
- [7][7] R. Tang and J. Lin, 'Deep residual learning for small-footprint keyword spotting,' in 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 2018, pp. 5484-5488. -
- [9][9] G. Chen, C. Parada, and T. N. Sainath, 'Query-by-example keyword spotting using long short-term memory networks,' in 2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) . IEEE, 2015, pp. 5236-5240. -
- [10][10] R. Menon, H. Kamper, E. van der Westhuizen, J. Quinn, and T. Niesler, 'Feature Exploration for Almost Zero-Resource ASRFree Keyword Spotting Using a Multilingual Bottleneck Extractor and Correspondence Autoencoders,' in Proc. Interspeech 2019 , 2019, pp. 3475-3479. [Online]. Available: http://dx.doi.org/10. 21437/Interspeech.2019-1665 -
- [13][13] T. Bluche and T. Gisselbrecht, 'Predicting Detection Filters for Small Footprint Open-Vocabulary Keyword Spotting,' in Proc. Interspeech 2020 , 2020, pp. 2552-2556. [Online]. Available: http://dx.doi.org/10.21437/Interspeech.2020-1186 - · DOI
- [14][14] A. Awasthi, K. Kilgour, and H. Rom, 'Teaching keyword spotters to spot new keywords with limited examples,' in Proc. Interspeech 2021 , 2021. -
- [17][17] R. David, J. Duke, A. Jain, V. J. Reddi, N. Jeffries, J. Li, N. Kreeger, I. Nappier, M. Natraj, S. Regev et al. , 'Tensorflow lite micro: Embedded machine learning on tinyml systems,' arXiv preprint arXiv:2010.08678 , 2020. - · arXiv
- [20][20] G. Klambauer, T. Unterthiner, A. Mayr, and S. Hochreiter, 'Selfnormalizing neural networks,' Advances in neural information processing systems , vol. 30, pp. 971-980, 2017. -
- [21][21] D. S. Park, W. Chan, Y. Zhang, C.-C. Chiu, B. Zoph, E. D. Cubuk, and Q. V. Le, 'SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition,' in Proc. Interspeech 2019 , 2019, pp. 2613-2617. [Online]. Available: http://dx.doi.org/10.21437/Interspeech.2019-2680 - · DOI · Paper page
- [23][23] E. Hermann and S. Goldwater, 'Multilingual bottleneck features for subword modeling in zero-resource languages,' in Proc. Interspeech 2018 , 2018, pp. 2668-2672. [Online]. Available: http://dx.doi.org/10.21437/Interspeech.2018-2334 - · DOI
Cited by
Sorted by citations
A review on TinyML: State-of-the-art and prospects
Ray, P. P. (2022) · Journal of King Saud University - Computer and Information Sciences · doi:10.1016/j.jksuci.2021.11.019 · Cited by 460
“A few-shot KWS is deployed in (Mazumder et al., 2021) that uses only 5 training examples which could be used for any language.”
Metadata
- Authors
- Mazumder, M., Banbury, C., Meyer, J., Warden, P., & Reddi, V. J.
- Affiliations
- Harvard University (US)
- Google (United States) (US)
- Venue
- Interspeech 2021
- Publisher
- ISCA
- Published
- 2021-08-30
- Pages
- 4214-4218
- ISSN
- 2308-457X (ISSN-L)
- OpenAlex
- W3144247233
- Cited by
- 38
- References
- 0
- Citations by year
- 2026: 4 · 2025: 11 · 2024: 5 · 2023: 10 · 2022: 4 · 2021: 4
- FWCI
- 2.66
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- Interspeech (USB) (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Speech Recognition and Synthesis
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Keywords
- keyword spotting, few-shot learning, multilingual speech recognition, transfer learning, embedding models, cross-lingual generalization, keyword search, false acceptance rate
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
