Acoustic model combination for recognition of speech in multiple languages using support vector machines | IEEE Conference Publication | IEEE Xplore