Few-Shot Keyword Spotting in Any Language
This enables efficient keyword spotting for low-resource languages, though it is incremental as it builds on existing transfer learning methods.
The authors tackled few-shot keyword spotting across any language by training a multilingual embedding model on nine languages and fine-tuning it with just five examples per keyword, achieving an average F1 score of 0.75 on 180 new keywords in those languages and 0.65 on 260 keywords in 13 new languages, with streaming accuracy of 87.4% at a 4.3% false acceptance rate.
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.