How to train your speaker embeddings extractor.
Mitchell McLarenDiego CastánMahesh Kumar NandwanaLuciana FerrerEmre YilmazPublished in: Odyssey (2018)
Keyphrases
- speaker verification
- speech recognition
- vector space
- audio visual
- speaker recognition
- dimensionality reduction
- low dimensional
- euclidean space
- machine learning
- artificial intelligence
- multi modal
- prosodic features
- manifold learning
- high dimensional data
- visual features
- evolutionary algorithm
- pattern recognition
- multiscale
- image sequences
- feature selection
- genetic algorithm