← all datasets

WSJ-0-2/3 Mix

Emerging
5papers using it
2021first seen

The 'WSJ-0-2/3 Mix' dataset is a benchmark used to evaluate speech separation performance, containing mixed speech signals from the Wall Street Journal corpus.

Papers using WSJ-0-2/3 Mix (5)

WSJ-0-2/3 Mix β€” datasets β€” speech-audio