Loading...
Please wait, while we are loading the content...
Similar Documents
Sparse Gaussian Graphical Models for Speech Recognition
| Content Provider | CiteSeerX |
|---|---|
| Author | Bell, Peter King, Simon |
| Abstract | We address the problem of learning the structure of Gaussian graphical models for use in automatic speech recognition, a means of controlling the form of the inverse covariance matrices of such systems. With particular focus on data sparsity issues, we implement a method for imposing graphical model structure on a Gaussian mixture system, using a convex optimisation technique to maximise a penalised likelihood expression. The results of initial experiments on a phone recognition task show a performance improvement over an equivalent full-covariance system. Index Terms: speech recognition, acoustic models, graphical models, precision matrix models |
| File Format | |
| Access Restriction | Open |
| Subject Keyword | Speech Recognition Sparse Gaussian Graphical Model Convex Optimisation Technique Equivalent Full-covariance System Automatic Speech Recognition Data Sparsity Issue Graphical Model Graphical Model Structure Initial Experiment Particular Focus Phone Recognition Task Inverse Covariance Matrix Acoustic Model Gaussian Graphical Model Penalised Likelihood Expression Gaussian Mixture System Performance Improvement Precision Matrix Model Index Term |
| Content Type | Text |
