Word-level segmentation

Hi everybody!

Our team is working now on the organization of the Third Workshop on NLP Applications to Field Linguistics. We are planning to launch a shared task aiming at automating word-level audio segmentation for field recordings of human speech made during fieldwork. ELAN is a tool which a lot of linguists. Therefore, we wonder in advance if it would be interesting for ELAN developers to intergrate the winner’s solution into ELAN or perhaps to take part in the workshop?

Hi!

Thanks for the announcement and the suggestion!

It’s difficult to say beforehand whether we can and want to integrate the winner’s solution; in general we are open to integrate code if it seems to be generic enough to be potentially useful for a considerable number of users. And there are technical preconditions of course (programming language, dependencies etc.).

Maybe a solution can be implemented as a Recognizer extension for which the original authors remain responsible? And ELAN can either include it in the distributions or advertise it as a third party component?

Hopefully it will be a successful workshop (not sure yet if we can take part)!

-Han

1 Like

Hi Han!

Thank you,

Maybe a solution can be implemented as a Recognizer extension for which the original authors remain responsible? And ELAN can either include it in the distributions or advertise it as a third party component?

yeah, a nice idea, we will consider it