Acoustic model merging using acoustic models from multilingual speakers for automatic speech recognition | IEEE Conference Publication | IEEE Xplore