Abstract
We investigate the efficiency of two very different spoken term detection approaches for transcription when the available data is insufficient to train a robust speech recognition system. This work is grounded in a very low-resource language documentation scenario where only a few minutes of recording have been transcribed for a given language so far. Experiments on two oral languages show that a pretrained universal phone recognizer, fine-tuned with only a few minutes of target language speech, can be used for spoken term detection through searches in phone confusion networks with a lexicon expressed as a finite state automaton. Experimental results show that a phone recognition based approach provides better overall performances than Dynamic Time Warping when working with clean data, and highlight the benefits of each methods for two types of speech corpus.
Original language | English |
---|---|
Title of host publication | Proceedings of the The 19th Annual Workshop of the Australasian Language Technology Association |
Editors | Afshin Rahimi, William Lane, Guido Zuccon |
Place of Publication | Australia |
Publisher | Australasian Language Technology Association |
Pages | 79-86 |
Number of pages | 8 |
Publication status | Published - Dec 2021 |
Event | 19th Workshop of the Australasian Language Technology Association, ALTA 2021 - Vitual, Online, Australia Duration: 8 Dec 2021 → 10 Dec 2021 |
Conference
Conference | 19th Workshop of the Australasian Language Technology Association, ALTA 2021 |
---|---|
Country/Territory | Australia |
City | Vitual, Online |
Period | 8/12/21 → 10/12/21 |