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: Manner of Articulation Detection using Connectionist Temporal Classification to Improve Automatic Speech Recognition Performance

Abstract: Conventionally, the manner of articulations in speech signal are derived using discriminative signal processing techniques or deep learning approaches. However, training such complex systems involves feature extraction, phoneme force alignment and deep neural network training. In our work, we initially detect the manner of articulations without phoneme alignment using an end-to-end manner of articulation modeling based on connectionist temporal classification (CTC). The manner of articulation knowledge is deployed in the conventional character CTC path to regenerate the new character CTC path. The modified manner based character CTC is evaluated on open source speech datasets such as AN4, LibriSpeech and TEDLIUM-2 and it outperforms over the baseline character CTC.
Comments: 5 pages, 4 figures, ICASSP-2019
Subjects: Audio and Speech Processing (eess.AS); Sound (cs.SD)
Cite as: arXiv:1811.01644 [eess.AS]
  (or arXiv:1811.01644v1 [eess.AS] for this version)

Submission history

From: Pradeep Rangan [view email]
[v1] Mon, 5 Nov 2018 12:43:05 GMT (189kb,D)

Link back to: arXiv, form interface, contact.