The SAIL Speaker Diarization System for Analysis of Spontaneous Meetings (bibtex)
by Han, Kyu J., Georgiou, Panayiotis G. and Narayanan, Shrikanth
Abstract:
In this paper, we propose a novel approach to speaker diarization of spontaneous meetings in our own mul- timodal SmartRoom environment. The proposed speaker di- arization system first applies a sequential clustering concept to segmentation of a given audio data source, and then performs agglomerative hierarchical clustering for speaker-specific classi- fication (or speaker clustering) of speech segments. The speaker clustering algorithm utilizes an incremental Gaussian mixture cluster modeling strategy, and a stopping point estimation method based on information change rate. Through experiments on various meeting conversation data of approximately 200 minutes total length, this system is demonstrated to provide diarization error rate of 18.90% on average.
Reference:
The SAIL Speaker Diarization System for Analysis of Spontaneous Meetings (Han, Kyu J., Georgiou, Panayiotis G. and Narayanan, Shrikanth), In Proceedings of IEEE International Workshop on Multimedia Signal Processing (MMSP), 2008.
Bibtex Entry:
@inproceedings{han_sail_2008,
	address = {Cairns, Australia},
	title = {The {SAIL} {Speaker} {Diarization} {System} for {Analysis} of {Spontaneous} {Meetings}},
	url = {http://ict.usc.edu/pubs/The%20SAIL%20Speaker%20Diarization%20System%20for%20Analysis%20of%20Spontaneous%20Meetings.pdf},
	abstract = {In this paper, we propose a novel approach to speaker diarization of spontaneous meetings in our own mul- timodal SmartRoom environment. The proposed speaker di- arization system first applies a sequential clustering concept to segmentation of a given audio data source, and then performs agglomerative hierarchical clustering for speaker-specific classi- fication (or speaker clustering) of speech segments. The speaker clustering algorithm utilizes an incremental Gaussian mixture cluster modeling strategy, and a stopping point estimation method based on information change rate. Through experiments on various meeting conversation data of approximately 200 minutes total length, this system is demonstrated to provide diarization error rate of 18.90\% on average.},
	booktitle = {Proceedings of {IEEE} {International} {Workshop} on {Multimedia} {Signal} {Processing} ({MMSP})},
	author = {Han, Kyu J. and Georgiou, Panayiotis G. and Narayanan, Shrikanth},
	month = oct,
	year = {2008}
}
Powered by bibtexbrowser