A Teacher-Student approach for extracting informative speaker embeddings from speech mixtures.
Tobias Cord-LandwehrChristoph BöddekerCatalin ZorilaRama DoddipatlaReinhold Haeb-UmbachPublished in: CoRR (2023)
Keyphrases
- speech recognition
- speaker recognition
- audio visual
- teacher student
- automatic speech recognition
- speaker verification
- speaker identification
- speaker dependent
- speaker diarization
- speech signal
- speech synthesis
- professional development
- online learning
- language model
- gaussian mixture model
- hidden markov models
- learning algorithm
- synthesized speech
- broadcast news
- data analysis
- case study