Code-switching in speech, the process of switching from one language to another
in the same conversational sequence, is common in bilingual and multilingual
communities and has been receiving growing attention in the speech and language
communities. The language switch can be just the incorporation of words or
short phrases from another language, or can be at a speaker turn or even larger
level. The video of our
workshop presentation describes our participation in the
shared task challenge organized by Microsoft Research on language identification
(LID) of code-switched speech in three language pairs: Gujarati-English,
Telugu-English and Tamil-English.
Vocapia, together with LIMSI-CNRS, have
been developing and participating in evaluations of LID technologies for more
than a decade. We have extensively explored a variety of approaches including
phonotactic and acoustic ones, e.g.
[Odyssey 2016] and
[InterSpeech
2016]. LIMSI also addressed Code-Switching in French/Algerian Arabic
Speech [Interspeech 2017].
Our submissions to the Microsoft challenge were ranked first in both LID
tasks: the first (task A) detecting if a given utterance was monolingual or
contained code-swithing; and the second (task B) ) producing a frame-level
language labeling of code-switched speech.
Vocapia Research, founded in July 2000, is an R&D company and
software publisher developing and providing leading edge speech
technologies and solutions for many languages, including most major
European Union languages as well as Arabic, Mandarin, and Russian. The
Vocapia Research VoxSigma
® software suite uses advanced
language technologies such as language identification, speech
recognition, and speaker identification to transform raw audio and
audiovisual data into structured and searchable XML documents. This
technology relies on over 25 years of research at LIMSI-CNRS, with
which there is a priviledged partnership. Joint systems developed
with LIMSI have achieved top ranks in national and international
challenges of speech-to-text transcription. The most common
applications of the VoxSigma software suite are audio and audiovisual
data mining (broadcast data, podcasts, call center data), media
monitoring, and media asset management. Vocapia Research is located in
the scientific pole of the Saclay Plateau, France. Readers who wish to
get more information about Vocapia Research are invited to check out
the Vocapia Research website or use the contact information page
http://www.vocapia.com/contact.