Skip to main content
Nasjonalbiblioteket

NoFA - Forced Alignment for Norwegian Bokmål

  • Datasets
  • Public access 

    Publicly available to everyone. Access may still require registration and an API key request, as long as anyone can request such registration and/or API keys.

    Read more about access levels here

  • Open data 

    The dataset is classified as public access and has at least one distribution with an approved open license.

Description

NoFA is a forced alignment model for Norwegian Bokmål, created by Nate Young (https://www.nateyoung.se/) for The Language Bank. This model is specifically made for the Montreal Forced Aligner (MFA), https://montreal-forced-aligner.readthedocs.io/.

Forced alignment refers to algorithms that take an audio file with speech and an orthographic transcription of the speech as input and produce a phonetic transcript where each segment (each phone) is time-aligned with the audio file.

NoFA is trained on The Language Bank's speech database NB Tale and the phonetic part of the RUNDKAST database developed at the Norwegian University of Science and Technology (NTNU).

See the documentation file for further information about NoFA.

Distributions
1

Nameless distribution
  • gtar
Description:
Not provided
Access URL:
https://hdl.handle.net/21.11146/59
Direct download:
API:
Not provided
Documentation:
Not provided
License:
Conforms to:
Not provided
Download

APIs providing this dataset
0

No registered APIs provide this dataset.

Similar datasets

NST Pronunciation Lexicon for SwedishNasjonalbiblioteket
Public access
Grapheme-to-Phoneme Models for NorwegianNasjonalbiblioteket
Public access
SCARRIE LexiconNasjonalbiblioteket
Public access
ONOMASTICA Pronunciation LexiconNasjonalbiblioteket
Public access
N-grams from NBdigital 2021Nasjonalbiblioteket
Public access