Free-Text Speaker Identification Over Long Distance Telephone Channel Using Hypothesized Phonetic Segmentation

Experimental investigation of free-text speaker identification method based on long-term statistics is conducted on a widely-used long distance telephone data base (l). On a 26-speaker subset, an average correct identification of 93.3% is obtained; on the complete 51speaker set, 67.6% correct identification is obtained. Speaker verification experiments on the data base provided receiver operating characteristics (ROC) comparable to or better than the ones available in open literature.