We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

eess.AS

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Electrical Engineering and Systems Science > Audio and Speech Processing

Title: QbyE-MLPMixer: Query-by-Example Open-Vocabulary Keyword Spotting using MLPMixer

Abstract: Current keyword spotting systems are typically trained with a large amount of pre-defined keywords. Recognizing keywords in an open-vocabulary setting is essential for personalizing smart device interaction. Towards this goal, we propose a pure MLP-based neural network that is based on MLPMixer - an MLP model architecture that effectively replaces the attention mechanism in Vision Transformers. We investigate different ways of adapting the MLPMixer architecture to the QbyE open-vocabulary keyword spotting task. Comparisons with the state-of-the-art RNN and CNN models show that our method achieves better performance in challenging situations (10dB and 6dB environments) on both the publicly available Hey-Snips dataset and a larger scale internal dataset with 400 speakers. Our proposed model also has a smaller number of parameters and MACs compared to the baseline models.
Comments: Accepted to INTERSPEECH 2022
Subjects: Audio and Speech Processing (eess.AS); Computation and Language (cs.CL); Machine Learning (cs.LG)
Cite as: arXiv:2206.13231 [eess.AS]
  (or arXiv:2206.13231v1 [eess.AS] for this version)

Submission history

From: Jinmiao Huang [view email]
[v1] Thu, 23 Jun 2022 18:18:44 GMT (653kb,D)

Link back to: arXiv, form interface, contact.