DialectID Save

Automatic Dialect Detection Repository

Project README

Automatic Arabic Dialect Detection Task

This code reflects the work described in the InterSpeech'2016 paper on Automatic Dialect Detection in Arabic Broadcast Speech.

It also contains a baseline system for the VarDial'2017 shared task on Arabic Dialect Identification.


Provided data:

  • We provide data for five Arabic dialects: Egyptian (EGY), Levantine (LAV), Gulf (GLF), North African (NOR), and Modern Standard Arabic (MSA).

  • The data comes from broadcast news.

VarDial'2017 shared task shared data, and features.

  • The baseline for VarDial'2017 is using data/train.vardial2017/ and data/dev.vardial2017/ for training and development default
  • For each dialect, there are two features files:
  • $dialect.words -- lexical features generated using LVCSR- generated using QCRI MGB-2 submission.
  • $dialect.ivec -- i-vector based on bottleneck features, with a fixed length of 400 per utterance.
  • wav.lst -- link to the original audio files; WAVE audio, Microsoft PCM, 16 bit, mono 16000 Hz.
  • Baseline-- bottleneck iVectors 57.28% accuracy and lexical features 48.43%.

InterSpeech'2016 paper shared data.

  • To reproduce the results in InterSpeech'2016, the script should point to data/train.IS2016/ and data/test.IS2016/ for training and testing.
  • $dialect.words -- lexical features generated using LVCSR;
  • $dialect.ivec -- i-vector based on bottleneck features, with a fixed length of 400 per utterance.
  • $dialect.phones -- phoneme sequence from an automatic phoneme recognition system.
  • $dialect.phone_duration -- phoneme sequence, and the duration in milliseconds for each phone, e.g., w_030 means phone w for 30 milliseconds.

Sample code

Run 'run.sh' for an example of the code and the data

  • features=phones -- you can use words, phones or ivectors;
  • context=6 -- for some features, less context might be enough;
  • NOTE 1: The regularization parameters can be optimized for better performance.
  • NOTE 2: System combination can be explored as well.


This data and the baseline system are described in this paper:

  author={Ali, Ahmed and Dehak, Najim and Cardinal, Patrick and Khurana, Sameer and Yella, Sree Harsha and Glass, James and Bell, Peter and Renals, Steve},
  title={Automatic Dialect Detection in Arabic Broadcast Speech},
  address={San Francisco, CA, USA}
Open Source Agenda is not affiliated with "DialectID" Project. README Source: qcri/dialectID
Open Issues
Last Commit
1 year ago

Open Source Agenda Badge

Open Source Agenda Rating