MCR-Data2vec 2.0: Improving Self-supervised Speech Pre-training via Model-level Consistency Regularization.
Ji Won YoonSeok Min KimNam Soo KimPublished in: INTERSPEECH (2023)
Keyphrases
- data sets
- experimental data
- input data
- prior knowledge
- data collection
- test data
- training samples
- measured data
- data structure
- probability distribution
- image data
- training data
- computational model
- simulation data
- data quality
- prior information
- mathematical model
- training examples
- database
- objective function
- data sources
- semi supervised learning
- missing data
- data processing
- probabilistic model
- statistical methods
- training set
- empirical data
- feature space
- data analysis