Skip to main content

Task 3 (under construction) : Unsupervised sound event clustering

Table of Contents

User scenario
#

The model benchmark produced in this task should provide evidence-based answers to the following user needs:

“What are the two thoughts popping up in my mind right after I got out of the water a new set of recordings ?

  • Is the quality of my data good enough to be exploited ?
  • What is in my data : anything interesting there ?

What kind of tools could I use to speed up the process of exploring new datasets and establishing exhaustive content reports able to inform on the presence of both targeted and noisy data sources ? “