Voxygen attends the Africacom exhibition

Voxygen attends AFRICACOM exhibition


ANR Half-time review on Nov 25th

The ANR Half-time review is planned on Nov 25th in Paris (ANR headquarters)

The slides presented can be found below


1st CFP of the 5th Workshop on Spoken Language Technologies for Under-resourced languages (SLTU 2016) in Indonesia

Call for Papers

The 5th International Workshop on Spoken Language Technologies for Under-resourced Languages (SLTU’16)

May 9-12, 2016, Yogyakarta, Indonesia

SLTU’16 Workshop web-site:  www.mica.edu.vn/sltu2016

### Important dates ###
– Full paper submission: 25 January 2016
– Notification of acceptance: 07 March, 2016
– Submission of final papers: 21 March, 2016
– Early registration: 21 March, 2016
– Workshop dates: 9-12 May, 2016

The Workshop on Spoken Language Technologies for Under-Resourced Languages is the fifth in a series of even-year SLTU Workshops. Four previous Workshops were organized: SLTU’14 in St. Petersburg (Russia), SLTU’12 in Cape Town (South Africa), SLTU’10 in Penang (Malaysia), and SLTU’08 in Hanoi (Vietnam).
SLTU’16 is held in Yogyakarta (Indonesia), a centre of classical Javanese culture. There are more than 700 ethnic languages spoken in Indonesian archipelago, but 146 are endangered. Therefore, SLTU’16 has special focus on under-resourced languages and endangered languages, but other related topics are also encouraged.

SLTU’16 will continue the tradition of previous SLTUs that features a number of distinguished keynote speaker, and this year, for the first time, SLTU will also offer Kaldi Tutorial on Under-Resourced Language.

### Workshop Topics ###
Areas related to processing any under-resourced and endangered  languages:
– Fast resources acquisition (text and speech corpora, parallel text, dictionary,
grammars, language model)
– Spoken language processing for language without dictionary or written forms
– Cross-lingual and multi-lingual spoken language processing including analysis
and synthesis
– Speech recognition and synthesis of low-resourced languages and dialects
– Machine translation and spoken dialogue systems
– Applications of spoken language technologies for under-resourced languages

### Paper submission ###
Prospective authors are invited to submit full-length papers up to 4 pages for technical content (including figures, tables, etc) and possible references, plus one additional (optional) page containing only references.

### Organizing Committee Members ###
– Workshop Chair: Sakriani Sakti (NAIST, Japan)
– Workshop Co-chair: Pascal Nocera (LIA, France), Eric Castelli (MICA, Vietnam), Laurent Besacier (LIG, France)
– Local Co-chair: Mirna Adriani (UI, Indonesia), Ayu Purwarianti (ITB, Indonesia)

Video about the multilingual calculator subtitled in English

Below, the video describing the multilingual calculator prototype developed by Voxygen.

In French, subtitled in English.


Paper accepted for the Show & Tell session at INTERSPEECH 2015

The Paper on Speech Technologies for African Languages: Example of a Multilingual Calculator for Education [PDF], wrote by Laurent Besacier, Elodie Gauthier, Mathieu Mangeot, Philippe Bretier, Paul Bagshaw, Olivier Rosec, Thierry Moudenc, Francois Pellegrino, Sylvie Voisin, Egidio Marsico, Pascal Nocera has been accepted to be presented at the Show & Tell session of INTERSPEECH 2015. The event will happen from September 6 to 10 and will be staged in the Maritim International Congress Center in Dresden (Germany).

This paper presents the achievements after 18 months of ALFFA project dealing with African languages technologies. It focus on a multilingual calculator prototype which will be demonstrated during the Show and Tell session.


Kaldi-based ASR system for Amharic now available to the public

All the data and scripts to build a complete ASR system for Amharic (based on Kaldi toolkit) is made available to the public on GitHub.

Amharic repository is composed of the transcribed speech corpus with audio, pronunciation lexicon and LM.

More details on Amharic ASR (read speech transcription, MER=8.43%) can be found on https://github.com/besacier/ALFFA_PUBLIC/blob/master/ASR/AMHARIC/README

LIG involved in BULB project (Breaking the Unwritten Language Barrier)


Capture d’écran 2015-03-17 à 14.27.28

The ANR/DFG project BULB (2015-2018) aims at supporting the documentation of unwritten languages with the help of automatic speech recognition and machine translation technologies. Three mostly unwritten Bantu languages will be at the heart of this project: Basaa, Myene and Embosi. The project relies on a strong German-French cooperation between linguists and computer scientists. The institutions involved together with ZAS (F. Hamlaoui) are the KIT (S. Stücker) and the University of Stuttgart (S. Zerbian) on the German side, as well as the LPP (M. Adda-Decker), the LLACAN (M. van de Velde), the LIMSI (L. Lamel and F. Yvon), the LIG (L. Besacier) and the IMMI-CNRS (G. Adda) on the French side

Swahili ASR system of LIG now included in the KALDI trunk

Swahili ASR system of LIG is now included in the KALDI trunk that can be downloaded by all KALDI users.

Voxygen presented its multilingual calculator at the Francophonie’s summit 2014 in Dakar

Voxygen presented its prototype of multilingual calculator to the “Village de la Francophonie” in Dakar from 24 to 30 November 2014.

See links below for more information:



The 21st Nov meeting notes are available on the intranet

The 21st Nov meeting notes are available on the intranet