Speaking rate dependent multiple acoustic models using continuous frame rate normalization | IEEE Conference Publication | IEEE Xplore