Finnish Dialect Identification: The Effect of Audio and Text
This work addresses the challenge of dialect identification for Finnish speakers, which is incremental as it applies existing multimodal methods to a new language context.
The paper tackled the problem of automatically identifying Finnish dialects from transcripts and audio recordings, achieving an accuracy of 85% when combining both modalities, compared to 57% with text alone.
Finnish is a language with multiple dialects that not only differ from each other in terms of accent (pronunciation) but also in terms of morphological forms and lexical choice. We present the first approach to automatically detect the dialect of a speaker based on a dialect transcript and transcript with audio recording in a dataset consisting of 23 different dialects. Our results show that the best accuracy is received by combining both of the modalities, as text only reaches to an overall accuracy of 57\%, where as text and audio reach to 85\%. Our code, models and data have been released openly on Github and Zenodo.