Self-supervised Fine-tuning for Improved Content Representations by Speaker-invariant Clustering.
Heng-Jui ChangAlexander H. LiuJames R. GlassPublished in: INTERSPEECH (2023)
Keyphrases
- fine tuning
- fine tuned
- invariant representations
- viable alternative
- clustering algorithm
- fine tune
- multiple representations
- clustering method
- data clustering
- metadata
- spectral clustering
- self organizing maps
- unsupervised learning
- speech recognition
- document clustering
- high dimensional data
- web documents
- hierarchical clustering
- affine transformation
- categorical data
- data points
- k means
- information theoretic
- fuzzy clustering
- image registration
- multimedia
- search engine