[DSTC11 Track3]

Skip to first unread message


Jul 12, 2022, 2:07:06 AMJul 12
to DSTC mailing list

Dear DSTC11 Track3 Organizers,

Hi, I have three questions.

  1. The DSTC track 3 proposal said that datasets would provide four formats (Human-Verbatim, Human-Paraphrased, Human-Noisy, and TTS-Verbatim). However, It seems provided version only contains TTS generated version. Has the task changed? Or the human data exists, and i can’t just find it?
  2. I understood that you’d given the data only for the first utterance of user speech. Am I right?
  3. In the TTS directory, there are 4 datasets (MUL, SUG, SSNG, WOZ). What is the difference between them? Can you give me more information about those datasets

Thank you in advance!
Reply all
Reply to author
0 new messages