WSJ-0-2mix-extr
Emerging4papers using it
2022first seen
The 'WSJ0-2mix-extr' dataset/benchmark contains mixed audio recordings of two speakers and is used to evaluate target-speaker automatic speech recognition (TS-ASR) performance.
Papers using WSJ-0-2mix-extr (4)
- Conformer-based Target-speaker Automatic Speech Recognition For Single-channel AudioSimultaneous Speech Extraction For Multiple Target Speakers Under The Meeting ScenariosSimultaneous Speech Extraction for Multiple Target Speakers under the
Meeting ScenariosConformer-based Target-Speaker Automatic Speech Recognition for
Single-Channel Audio