Research project ASR-UD: Automatic Speech Recognition for Upper Dalecarlian
ASR-UD develops language-technology infrastructure and digital tools for the Upper Dalecarlian varieties (Moramål, Oremål, Orsamål, Sollerömål, Venjansmål, Våmhusmål and Övdalian/Elfdalian). The project includes automatic transcription, phonetic annotation and, in the longer term, other forms of automatic linguistic analysis.
The Upper Dalecarlian varieties (Moramål, Oremål, Orsamål, Sollerömål, Venjansmål, Våmhusmål and Övdalian/Elfdalian) have limited access to digital language resources and tools for processing speech and text. ASR-UD develops language-technology infrastructure for research on these varieties, with automatic transcription as an initial focus.
The work includes corpus construction, transcription normalization, speech recognition, and the handling of linguistic and orthographic variation. The infrastructure is being expanded to support additional tasks, including phonetic annotation and automatic morphosyntactic analysis such as part-of-speech tagging.
The aim is to make larger quantities of speech and text data accessible for linguistic analysis while developing reusable digital resources for research on Upper Dalecarlian.
Visual concept developed by Nathan Young using ChatGPT 5.2 (OpenAI).