ISCA & IEEE Workshop on Spontaneous Speech Processing and Recognition
April 13-16, 2003
We describe a recognition experiment and two analytic experiments on a database of strongly Hispanic-accented English. We show the crucial importance of training on the Hispanicaccented data for acoustic model performance, and describe the tendency of Spanish-accented speakers to use longer, and presumably less-reduced, schwa vowels than native-English speakers.
Bibliographic reference. Ikeno, Ayako / Pellom, Bryan / Cer, Dan / Thornton, Ashley / Brenier, Jason M. / Jurafsky, Dan / Ward, Wayne / Byrne, William (2003): "Issues in recognition of Spanish-accented spontaneous English", in SSPR-2003, paper MAP7.