{"date_updated":"2022-01-06T06:51:08Z","type":"conference","publication":"ICASSP99 Phoenix, AZ","main_file_link":[{"open_access":"1","url":"https://groups.uni-paderborn.de/nt/pubs/1999/ICASSP_1999_Haeb_paper.pdf"}],"language":[{"iso":"eng"}],"title":"Investigations on inter-speaker variability in the feature space","_id":"11780","department":[{"_id":"54"}],"date_created":"2019-07-12T05:28:07Z","user_id":"44006","citation":{"bibtex":"@inproceedings{Haeb-Umbach_1999, title={Investigations on inter-speaker variability in the feature space}, booktitle={ICASSP99 Phoenix, AZ}, author={Haeb-Umbach, Reinhold}, year={1999} }","short":"R. Haeb-Umbach, in: ICASSP99 Phoenix, AZ, 1999.","chicago":"Haeb-Umbach, Reinhold. “Investigations on Inter-Speaker Variability in the Feature Space.” In <i>ICASSP99 Phoenix, AZ</i>, 1999.","ieee":"R. Haeb-Umbach, “Investigations on inter-speaker variability in the feature space,” in <i>ICASSP99 Phoenix, AZ</i>, 1999.","ama":"Haeb-Umbach R. Investigations on inter-speaker variability in the feature space. In: <i>ICASSP99 Phoenix, AZ</i>. ; 1999.","mla":"Haeb-Umbach, Reinhold. “Investigations on Inter-Speaker Variability in the Feature Space.” <i>ICASSP99 Phoenix, AZ</i>, 1999.","apa":"Haeb-Umbach, R. (1999). Investigations on inter-speaker variability in the feature space. In <i>ICASSP99 Phoenix, AZ</i>."},"year":"1999","status":"public","author":[{"id":"242","last_name":"Haeb-Umbach","full_name":"Haeb-Umbach, Reinhold","first_name":"Reinhold"}],"oa":"1","abstract":[{"lang":"eng","text":"We apply Fisher variate analysis to measure the effectiveness of speaker normalization techniques. A trace criterion, which measures the ratio of the variations due to different phonemes compared to variations due to different speakers, serves as a first assessment of a feature set without the need for recognition experiments. By using this measure and by recognition experiments we demonstrate that cepstral mean normalization also has a speaker normalization effect, in addition to the well-known channel normalization effect. Similarly vocal tract normalization (VTN) is shown to remove inter-speaker variability. For VTN we show that normalization on a per sentence basis performs better than normalization on a per speaker basis. Recognition results are given on Wall Street Journal and Hub-4 databases"}]}