Graph-PIT: Generalized Permutation Invariant Training for Continuous Separation of Arbitrary Numbers of Speakers.
Thilo von NeumannKeisuke KinoshitaChristoph BöddekerMarc DelcroixReinhold Haeb-UmbachPublished in: Interspeech (2021)
Keyphrases
- speech recognition
- graph structure
- training set
- weighted graph
- training samples
- training algorithm
- random walk
- graph construction
- directed graph
- graph representation
- graph partitioning
- graph model
- spanning tree
- training process
- affine transformation
- supervised learning
- social networks
- training phase
- discrete valued
- minimum classification error
- connected graphs
- directed acyclic graph
- affine invariant
- graph theory
- bipartite graph
- connected components
- artificial neural networks
- object recognition
- bayesian networks