Subspace Gaussian mixture model explained

Subspace Gaussian mixture model (SGMM) is an acoustic modeling approach in which all phonetic states share a common Gaussian mixture model structure, and the means and mixture weights vary in a subspace of the total parameter space.^[1]

Notes and References

1. Povey, D : Burget, L.; Agarwal, M.; Akyazi, P. "Subspace Gaussian Mixture Models for speech recognition", IEEE, 2010, Acoustics Speech and Signal Processing (ICASSP), 2010 IEEE International Conference on, pp. 4330–33, doi:10.1109/ICASSP.2010.5495662