German Dialects: Document, Preserve and Learn (with the help of AI)
Project Description
The project is being realised in collaboration with Ludwig Maximilian University of Munich, the Austrian Research Institute for Artificial Intelligence, and the University of Liechtenstein. The objective is to document and study the dialects of Liechtenstein, western Austria, and South Tyrol.
The project is divided into three phases:
Phase 1: Documentation and archiving: Collection of audio recordings in dialects from public sources, enriched with audio recordings created specifically for the project.
Phase 2: Machine processing of the audio collection: The collected audio files are automatically transcribed into Standard German and metadata is created: gender of the speaker, associated dialect region and age group.
Phase 3: Learning: The used AI models are offered to interested parties on a platform. Instructions are created for linguists and AI researchers on how these models can be integrated into their research work. The platform also supports the learning of dialects. In this way, it helps to overcome the language barrier for relocators.
The project is divided into three phases:
Phase 1: Documentation and archiving: Collection of audio recordings in dialects from public sources, enriched with audio recordings created specifically for the project.
Phase 2: Machine processing of the audio collection: The collected audio files are automatically transcribed into Standard German and metadata is created: gender of the speaker, associated dialect region and age group.
Phase 3: Learning: The used AI models are offered to interested parties on a platform. Instructions are created for linguists and AI researchers on how these models can be integrated into their research work. The platform also supports the learning of dialects. In this way, it helps to overcome the language barrier for relocators.
