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

Download:

Current browse context:

cs.CL

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

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

Computer Science > Computation and Language

Title: Instant One-Shot Word-Learning for Context-Specific Neural Sequence-to-Sequence Speech Recognition

Abstract: Neural sequence-to-sequence systems deliver state-of-the-art performance for automatic speech recognition (ASR). When using appropriate modeling units, e.g., byte-pair encoded characters, these systems are in principal open vocabulary systems. In practice, however, they often fail to recognize words not seen during training, e.g., named entities, numbers or technical terms. To alleviate this problem we supplement an end-to-end ASR system with a word/phrase memory and a mechanism to access this memory to recognize the words and phrases correctly. After the training of the ASR system, and when it has already been deployed, a relevant word can be added or subtracted instantly without the need for further training. In this paper we demonstrate that through this mechanism our system is able to recognize more than 85% of newly added words that it previously failed to recognize compared to a strong baseline.
Comments: 7 pages, 1 figure, 4 tables
Subjects: Computation and Language (cs.CL); Machine Learning (cs.LG)
Cite as: arXiv:2107.02268 [cs.CL]
  (or arXiv:2107.02268v1 [cs.CL] for this version)

Submission history

From: Christian Huber [view email]
[v1] Mon, 5 Jul 2021 21:08:34 GMT (127kb,D)

Link back to: arXiv, form interface, contact.