Comparing one and two-stage acoustic modeling in the recognition of emotion in speech | IEEE Conference Publication | IEEE Xplore