Speaker identification using higher order spectral phase features and their effectiveness vis-a-vis Mel-Cepstral features

The effectiveness of higher-order spectral (HOS) phase features in speaker recognition is investigated by comparison with Mel Cepstral features on the same speech data. HOS phase features retain phase information from the Fourier spectrum unlikeMel–frequency Cepstral coefficients (MFCC). Gaussian mixture models are constructed from Mel– Cepstral features and HOS features, respectively, for the same data from various speakers in the Switchboard telephone Speech Corpus. Feature clusters, model parameters and classification performance are analyzed. HOS phase features on their own provide a correct identification rate of about 97% on the chosen subset of the corpus. This is the same level of accuracy as provided by MFCCs. Cluster plots and model parameters are compared to show that HOS phase features can provide complementary information to better discriminate between speakers.

Perpetuating Racism: Legal Education, Race and Social Justice

By presenting the results of a content analysis of Australian undergraduate legal education, this paper examines the extent to which issues of race, ethnicity, discrimination, and multiculturalism feature within this component of the moral, ethical, and professional development of legal professionals. It will demonstrate that instead of encouraging a deep, critical and contextual understanding of such issues, legal education provides a relatively superficial one, which has important implications for the role that legal professionals play in overcoming injustices such as institutional racism, and the kinds of social reform that they are likely to undertake.