Swiss-German-Speech-Data---Meeting-Notes.md

@@ -22,9 +22,8 @@ Yixuan Xu, Daniel Perruchoud, Michael Graber
     * Synthetic Swiss German data?
   * Daniel: Idiotikon data might be an option
   * Permissive licence
+* Ideally for stage 2 would be daily conversations, no bias towards political parliament data
 * SRF -\> Ablation study could be done (\~ 2000 h)
-
-
 * FHNW: Transcription Evaluation for SRF broadcast transcripts
   * FHNW models perform substantially better than current SRF solution
   * ![TranscriptionRatingDistributionSRFvsFHNW.png](uploads/f4f0cc084aa9daae243cf90cb83e3485/TranscriptionRatingDistributionSRFvsFHNW.png){width="504" height="360"}