SITW
Emerging5papers using it
2021first seen
The SITW dataset is used to evaluate speaker recognition systems by providing a collection of speech samples that facilitate the assessment of speaker trait and content disentanglement.
Papers using SITW (5)
- Incorporating Uncertainty from Speaker Embedding Estimation to Speaker
VerificationDisentangling Voice and Content with Self-Supervision for Speaker
RecognitionAttention-based conditioning methods using variable frame rate for
style-robust speaker verificationTask-aware Warping Factors in Mask-based Speech EnhancementOptimizing Multi-Taper Features for Deep Speaker Verification