Tytuł artykułu

An Effective Speaker Clustering Method using UBMand Ultra-Short Training Utterances

Tytuł czasopisma

Archives of Acoustics




No 1

Autorzy publikacji

Słowa kluczowe

automatic speech recognition; interindividual difference compensation; speaker clustering; universal background model; GMM weighting factor adaptation

Wydział PAN

Nauki Techniczne


The same speech sounds (phones) produced by different speakers can sometimes exhibit significant differences. Therefore, it is essential to use algorithms compensating these differences in ASR systems. Speaker clustering is an attractive solution to the compensation problem, as it does not require long utterances or high computational effort at the recognition stage. The report proposes a clustering method based solely on adaptation of UBM model weights. This solution has turned out to be effective even when using a very short utterance. The obtained improvement of frame recognition quality measured by means of frame error rate is over 5%. It is noteworthy that this improvement concerns all vowels, even though the clustering discussed in this report was based only on the phoneme a. This indicates a strong correlation between the articulation of different vowels, which is probably related to the size of the vocal tract.


Committee on Acoustics PAS, PAS Institute of Fundamental Technological Research, Polish Acoustical Society


ISSN 0137-5075 ; eISSN 2300-262X