Swiss German Data Collection Tool
As a part of a bachelor thesis, we developed a platform which gathers text and audio samples of Swiss German speakers, in order to support the development of Swiss German NLP applications. We extended the Voice Collection tool by the Fachhochschule Nordwestschweiz (FHNW) with our new features. The platform supports, amongst other applications, the following use cases:
- The creation of a translation corpus for German text to Swiss German text.
- The creation of a text-speech corpus for Swiss German text to Swiss German speech.
- The creation of spontaneous text and speech recordings by visual inspirations.
All of the above mentioned data collection options are enhanced by gamification elements to keep users motivated and provide feedback considering the data quality.
The code can be found here: external pagehttps://github.com/cintoros/speech-collection-app
The platform can be tested here: external pagehttps://www.cs.technik.fhnw.ch/speech-to-text-labeling-tool/home